WRITTEN STŪPA, PAINTED SŪTRA: RELATIONSHIPS OF TEXT AND IMAGE IN THE

CONSTRUCTION OF MEANING IN THE JAPANESE JEWELED-STŪPA MANDALAS

BY
©2011
HALLE ELIZABETH O‘NEAL

Submitted to the graduate degree program in The Kress Foundation Department of Art History
and the Graduate Faculty of the University of Kansas in partial fulfillment of the requirements
for the degree of Doctor of Philosophy.

________________________________
Chairperson Sherry Fowler

________________________________
Marsha Haufler

________________________________
Amy McNair

________________________________
Sally Cornelison

________________________________
Margaret Childs

Date Defended: October 19, 2011

 The Dissertation Committee for HALLE ELIZABETH O‘NEAL
certifies that this is the approved version of the following dissertation:

WRITTEN STŪPA, PAINTED SŪTRA: RELATIONSHIPS OF TEXT AND IMAGE IN THE

CONSTRUCTION OF MEANING IN THE JAPANESE JEWELED-STŪPA MANDALAS

________________________________
Chairperson Sherry Fowler

Date approved: October 19, 2011

ii
 Abstract

This dissertation contextualizes the twelfth- and thirteenth-centuries Japanese jeweled-stūpa

mandalas as some of the most striking examples from the early medieval period of innovative

elaborations on sūtra transcription. The project proceeds from a methodology grounded in visual

analysis and religious studies. I begin with basic questions of semiotic inquiry about the

prominence and privileging of sacred text in the form of the central dharma reliquary, a

characteristic distinguishing the mandalas from nearly all other paintings made before them. I

seek to understand the reasons behind the privileging of scripture on the picture plane and the

inventive manipulation of the sūtra text into the form of a stūpa, both novel choices in the

context of their early medieval Japanese production.

At their root, the jeweled-stūpa mandalas are an elaborate sūtra transcription project

revealing anxieties about death and power expressed through the belief that devotion to sūtra can

save souls, cure illnesses, grant tremendous authority, and much more. After investigating the

continental origins of the mandalas and the culture of sūtra transcription during the eleventh

through thirteenth centuries and conducting an analysis into the particular histories and formal

qualities, the project approaches the mandalas using a three-part collaborative analysis. The first

part examines visual, textual, and archaeological evidence from the eleventh through thirteenth

centuries, which testifies to the understandings and capabilities of text as well as the power of

sacred word expressed repeatedly and profoundly in early medieval Japan. This exploration of

sūtra text lays the critical basis for the second part‘s investigation into the notion of body

underpinning the innovative construction of the mandalas. The indivisibility of sūtra, stūpa,

iii
dharma, relic, and body in the paintings visually manifests the conflated nature of these

seemingly independent concepts in religious practice and doctrine. Combining the first two parts

facilitates a reading of the mandalas through what I call a salvific matrix of text and body. The

third part concludes the dissertation by returning to an explicit discussion of semiotics, further

exploring the construction of meaning in the mandalas through their imbrication of text and

image.

iv
 Dedication

For Jerry O‘Neal,

Entomologist, poet, and author of historical westerns—but most important to me,

loving father and the most compassionate of men.

I miss you and love you so.

v
 Acknowledgements

This project began to take shape in 2005 during a graduate seminar about the visual culture of

relics and reliquaries in Buddhism and Christianity. In addition to her influence as co-convener

of that seminar, I owe the greatest debt of gratitude to my advisor, Sherry Fowler, not only

because she recognized even then the worth of the jeweled-stūpa mandalas as a dissertation topic,

but also for her faithful guidance, support, and encouragement. Sherry is always generous with

her time and judgment, and her instruction throughout my graduate studies has been instrumental

to my thinking about Asian art history in general and to this project in particular. I would also

like to warmly thank the other members of my dissertation committee, Amy McNair, Marsha

Haufler, Sally Cornelison, and Maggie Childs, for their discerning comments and suggestions.

This dissertation is stronger because of their participation in the project, and any remaining

weaknesses and missteps are purely my own responsibility. I would especially like to thank Amy

McNair for being the second-reader and for her insightful instruction in the classroom, and

Marsha Haufler for her inspirational courses which kindled my secondary love of Chinese

painting.

Donohashi Akio of Kōbe University, Japan, gave freely of his time and extensive

knowledge of Buddhist painting during my research period there and skillfully arranged for me

to examine the Danzan Shrine jeweled-stūpa mandalas. Because of his generous spirit and the

dear friends I made there, I felt completely welcomed into the art history department at Kobe.

Additionally, I thank Naitō Sakae of Nara National Museum for allowing me to attend his highly

informative lectures on Buddhist visual and material culture. I have also benefited greatly from

vi
presenting different aspects of this project at several conferences, where I received valuable

feedback from many individuals, in particular Ian Astley, Michael Jamentz, Lucia Dolce, and

John Carpenter. Oikawa Kaori helped me slog through the translation of passages that were

critical to the project. The many conversations with friends in the art history cohort of both the

University of Kansas and Kōbe University helped me clarify my project and retain my sanity.

Throughout my graduate career, I received generous funding for both dissertation

research and language training. The Japanese government granted an eighteen-month

Monbukagakushō Scholarship which supported my time at Kōbe University. I was also

substantially funded by Foreign Language and Area Studies fellowships through the Center for

East Asian Studies at the University of Kansas in the summers of 2003, 2004, and 2009 and in

the academic years of 2003-2004 and 2006-2007. The University of Kansas funded the

presentation of my research in 2007 with a Graduate Student Paper Presenter Travel Fund and an

Art History Travel Fund. The US State Department supported further study of advanced modern

Japanese in summer 2010 through a Critical Language Scholarship. I would not have been able

to complete the dissertation without these various forms of support and am grateful for all of

them.

My debt to family and friends cannot be fully acknowledged here without significant

space and embarrassment. My mother, Sylvia O‘Neal, provided support and love and kept me

well-fed and laughing throughout the process. My mother-in-law, Susan Hom, provided loving

attentiveness at various stages and is generally the most considerate person I‘ve ever had the

pleasure of knowing. My father-in-law, Harry Hom, provided the constant model of a fun-loving

yet engaged academic. My aunt, Brenda Lott, has been a great source of support and

encouragement throughout my entire life. To my husband, Andy Hom, who brings utter

vii
happiness and love to my life, I thank you for the years of unfailing support and comfort and for

reading numerous drafts of this dissertation and countless term papers during graduate school.

Our years together have been the best of my life, and I look forward with eager anticipation to

our ‗herd.‘ Finally, I would like to thank my dedicated father, Jerry O‘Neal aka Jess McCreede,

for raising me and for always loving me. The passage of time cannot diminish my devotion to

you.

viii
 Table of Contents

Chapter One: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Introduction to Topic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
Review of the Jeweled-Stūpa Mandala Literature . . . . . . . . . . . . . . . . . . .7
Plan of the dissertation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Chapter Two: Continental Prototypes and Contexts of Copying . . . . . . . . . . . . 17

Continental Prototypes 18
Culture of Copying 26
Conclusion 51

Chapter Three: The Jeweled-Stūpa Mandalas as Historical Objects: Analyses

of the Formal Qualities and Context . . . . . . . . . . . . . . . . . . . . . . . 52
Set History: Chūsonji Mandalas 52
Set History: Danzan Shrine Mandalas 85
Set History: Ryūhonji mandalas 105
Lone Jeweled-Stūpa Mandalas 118
Conclusion 123

Chapter Four: Dharma Reliquary and the Power of Word . . . . . . . . . . . . . . 125

Power of Sacred Word .126
Dharma Relics 146
Conclusion 174

Chapter Five: Jeweled-Stūpa Mandalas as a Salvific Matrix of Text and Body . . . . 176
Bodies of the Buddha 178
Choosing the Stūpa 188
Conclusion: the Salvific Matrix of Text and Body 205

Chapter Six: Text and Image Issues in the Jeweled-Stūpa Mandalas . . . . . . . . . 213
Semiotic Perspectives 214
Textualized Community 218
Role Reversals 240
Ways of Viewing 250
Materiality 253
Conclusion 260

Chapter Seven: Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262

Goals and Summary of the Project 262
Avenues for Future Research 277

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

ix
 Chapter One

Introduction

Introduction to Topic

Radiant gold and traces of oxidized silver contrast dramatically against a deep blue background

in the Japanese jeweled-stūpa mandalas (Kinji hōtō mandara 金字宝塔曼荼羅) of the twelfth

and thirteenth centuries. Tall and narrow, the format complements the structure of the central,

golden stūpa. Rooted in the earthly realm, yet existing in an otherworldly space, the stūpa is

surrounded by graphic vignettes adapted from the tales of the sūtras. Rather than straight and

measured architectural lines, diminutive sūtra characters build and fill the body of the stūpa.

Painstakingly constructed from one of two popular and potent scriptures, the Lotus Sūtra1or the

Golden Light Sūtra,2 each mandala set produces a particular and complete scripture in the form

of textual reliquaries.

This project focuses on the mandala sets from Chūsonji 中尊寺 in Hiraizumi, Danzan

Shrine 談山神社 in Nara, and Ryūhonji 立本寺 in Kyoto, along with two other mandalas

separated from their original sets. Chūsonji‘s set of ten mandalas are visual translations of the

Golden Light Sūtra and were likely commissioned around 1170 by Fujiwara Hidehira 藤原秀衡

(1122-1187). The Danzan Shrine version translates the beloved Lotus Sūtra into the jeweled-

stūpa mandala format, but with the addition of two bracketing scriptures—the Innumerable

1
Jpn. Myōhō renge kyō; Ch. Miaofa lianhua jing; Skt. Saddharmapuṇḍarīka sūtra; 妙法蓮華経; Taishō daizōkyō
大正大藏経, ed. Takakusu Junjirō 高楠順次郎 and Watanabe Kaigyoku 渡邊海旭 (Tokyo: Taishō Issaikyō
Kankōkai, 1924-32) no. 262, 9: 1c15-62b1. Hereafter abbreviated as T. Texts are indicated by the text number
followed by the volume, page, register, and line numbers, where appropriate.
2
Jpn. Konkōmyō saishōō kyō; Ch. Jinguangming zuisheng wang jing; Skt. Suvarṇaprabhāsottama rāja sūtra; 金光
明最勝王経; T. no. 665, 16: 403a04-456c25.

1
Meanings Sūtra3as the prologue and Sūtra of Meditation on the Bodhisattva Universal Worthy4

as the epilogue—to form a set of ten mandalas dating from the twelfth century. Ryūhonji‘s

jeweled-stūpa mandalas of the early thirteenth century also capture the twenty-eight chapters of

the Lotus Sūtra, both textually and narratively, in eight mandalas.

Paintings inspired by scripture have been widely studied in the Japanese art historical

field, yet paintings composed of or dominated by textual dharma—the written teachings of the

Buddha venerated as relics—have received far less attention. However, this marginalization

within scholarship may not accurately reflect the power of word in early medieval 5 Japanese

religious and cultural society. The paintings I examine in this project exhibit a reverent and

inventive use of religious text, manifesting the belief that words have salvific potency. Thus, this

dissertation begins with ostensibly simple questions: What is the purpose of featuring Buddhist

scriptural text prominently in the paintings? Why does textual dharma replace image, or more

specifically, become image? What do these relationships suggest about the power of word and

language? These general queries contain densely packed interdisciplinary issues regarding the

power of written word as well as broader discussions about the relationship between text and

image.

A recent symposium publication on texts and writing in premodern Japan begins with the

following statement:

3
Jpn. Muryōgi kyō; Ch. Wuliangyi jing; Skt. Amitartha sūtra; 無量義経; T. no. 276, 9: 383b15-389b22.
4
Jpn. Kan fugen bosatsu gyōhō kyō; Ch. Guan puxian pusa xingfa jing; 觀普賢菩薩行法経; T. no. 277, 9: 389b26-
394b11.
5
I have chosen to use the term ‗early medieval‘ to refer to the ninth through thirteenth centuries. While it is a loaded
term and one with many definitions, including those that disagree with this particular dating, it is also a concise and
handy designation. These centuries roughly correspond to the Heian period (794-1185) and most of the Kamakura
period (1185-1333), and the term is meant to suggest some consistency in the foundations of art and Buddhism
during this time. When the term ‗medieval‘ is used, I refer to the ninth through sixteenth centuries, the most
accepted closing date of the medieval period. Often, however, I explicitly refer to the eleventh through thirteenth
centuries.

2
 Written sources of various kinds, official and unofficial, have everywhere been the
marrow of history. But writing is not the only medium through which accounts of the past
have been expressed; texts can be transmitted orally, represented visually, performed on
stage, or preserved and passed on in other ways. And once a text exists, it is available for
reference, recycling, revision—and a number of uses and abuses—by those who
encounter it.6

Texts can be understood and manifested in a variety of ways. A central theme of this dissertation

concerns the various lives of sacred texts examined through an art historical lens. Fabio Rambelli

posits that medieval texts had more than ―just a ‗meaning‘—understood as the ‗signified‘ of the

text itself as the ‗signifier.‘‖ 7 Medieval Japanese texts, and indeed the characters which

composed them, led multifaceted and ever-changing lives which often exceeded the bounds of

the staple semiotic equation of signified and signifier. Thus, it is the textual web, which is

composed of various lives lived, artistic permutations, and the power of scriptural word, that

forms the basis of my investigation.

At their root, the jeweled-stūpa mandalas are an elaborate sūtra transcription project

revealing anxieties about death and power expressed through the belief that devotion to sūtra text

can save souls, cure illnesses, grant tremendous authority, and much more. After investigating

the continental origins of the mandalas and the culture of sūtra transcription during the eleventh

through thirteenth centuries in chapter two and conducting an analysis into the particular

histories and formal qualities in chapter three, the project approaches the mandalas using a three-

part collaborative analysis. The first part examines visual, textual, and archaeological evidence

from the eleventh through thirteenth centuries, which testifies to the understandings and

capabilities of text as well as the power of sacred word expressed repeatedly and profoundly in

6
James C. Baxter and Joshua A. Fogel, eds., Writing Histories in Japan: Texts and Their Transformations from
Ancient Times through the Meiji Era (Kyoto: International Research Center for Japanese Studies, 2007), 1.
7
Fabio Rambelli, ―Texts, Talismans, and Jewels: The Reikiki and the Performativity of Sacred Texts in Medieval
Japan,‖ in Discourse and Ideology in Medieval Japanese Buddhism, ed. Richard Karl Payne and Taigen Dan
Leighton (New York: Routledge, 2006), 52.

3
early medieval Japan in the fourth chapter. This exploration of sūtra text lays the critical basis for

the second part‘s further investigation into the notion of body underpinning the innovative

construction of the mandalas in chapter five. The indivisibility of sūtra, stūpa, relic, and body in

the paintings visually manifests the conflated nature of these seemingly independent concepts in

religious practice and doctrine. Chapters four and five culminate in a reading of the mandalas

through what I call a salvific matrix of text and body.

Although the overall project is grounded in the question concerning the construction of

meaning and reality as portrayed in the jeweled-stūpa mandalas, the third part specifically

addresses the visual relationship of text and image and the role of word in a medium normally

dominated by picture through a semiotic perspective. This analysis addresses the inversion of the

roles experienced by text and image and the subsequent complication of the conventional

semiotic relationship between signifier and signified as well as the orality and performative

aspects of texts in early medieval Japan. Notions of text and image are fluid and permeable in the

jeweled-stūpa mandalas. These malleable roles foster compelling questions about the nature of

text and image in Japanese Buddhist painting, thus fashioning an intriguing art historical puzzle.

What are the visual functions of written word in a painting where word cannot be read? As a

textually imbricated image, how can or should we view the dharma reliquary? What shifts in the

conventional functions of text and image occur when word becomes picture and pictures tell the

stories word no longer can, as in the case of the narrative vignettes along the sides of the

mandalas? In what ways are the definitions of text and image challenged or inverted by such

collusions? These role reversals, complications of conventional functions, and layered viewing

issues constitute the primarily visual thrust of my analysis of the jeweled-stūpa mandalas.

4
Lives of Texts

Language cannot be decontextualized. Any examination into the role of textual dharma must be

careful not to strip texts of their context and deprive them of their full voice and place in the

historical milieu, which would commit the error of privileging and isolating doctrine apart from

understandings of praxis. In this regard, I focus on the visual manifestations of scripture and on

revealing the different lives of those sacred texts and their implications for soteriological and

apotropaic power invested in word.

Offertory reading is the oral performance of texts as ritual worship and does not oblige

hermeneutical depth, whereas interpretative reading is the exegetical examination of texts for

their substantive meaning.8 While a plethora of early medieval written sources confirm this dual

scheme, I would propose here an additional concept of visual or artistic reading. Optical

registering of graphic images suggests a visual mode of reading—one in which the viewer

processes the graphic image for its interconnected parts. This is particularly true in the case of

jeweled-stūpa mandalas where the image is composed of or dominated by text, compelling the

viewer to not only read the graphic components of the painting but also the textual ones. Such

visual reading of paintings infused with dharma relics allows for the cognizance of sacred word‘s

power and suggests that early medieval religious society embraced a more nuanced

understanding of artistic ‗literacy‘ than has been previously explored in the literature.

Certainly, texts were valued beyond their discursive function for their performative

qualities and for their material expression of the immaterial. This physical expression constituted

various systems of value, from economic to symbolic and religious currency. 9 Indeed the

hermeneutical sense of reading was not the primary purpose of sacred texts, for the vast and

8
Thomas R. Howell, ―Setsuwa, Knowledge, and the Cultures of Reading and Writing in Medieval Japan‖ (PhD diss.,
University of Pennsylvania, 2002), 172.
9
Rambelli, ―Texts, Talismans, and Jewels,‖ 52-53.

5
influential meanings of sacred words extended far beyond what was signified. 10 Especially in

esoteric forms of Buddhism, the orality of medieval textuality played a crucial role in textual

function. ―Reading was usually not silent, but voiced‖ in medieval Japan. 11 The consumption of

sacred texts, such as kuden 口伝 (the oral transmission of specific teachings from master to

disciple in the Tendai tradition), kanjin 観心 (the ‗contemplation of the mind,‘ a meditative

interpretation of sacred texts), and kanjō 潅頂 (in part involving the secret transmission of texts)

suggests the various ingestions and performances of malleable, medieval texts.

The countless explications and manifestations of sacred word in art, literature, and poetry

suggest that scriptures are open texts, capable of potentially endless re-creation and

reinterpretation, 12 which the Shingon 真言 monk, Kūkai 空海 (774-835), claimed necessitated

constant and pious re-construction. Ryūichi Abé explains: ―Kūkai approaches the text as a yet-to-

be bound—or, perhaps more appropriately, never-to-be bound—constantly reworked manuscript.

For Kūkai, the text is not a book but a writing that remains open-ended.‖13 The many and

inventive artistic permutations of sacred word not only illustrate the concept of open texts, but

also embody and manifest the great power of textual dharma. It is this power of dharma relics,

both salvific and restorative, that perhaps compels its manifestation in visual culture.

10
Fabio Rambelli, Buddhist Materiality: A Cultural History of Objects in Japanese Buddhism (Stanford: Stanford
University Press, 2007), 88-90.
11
Ibid., 94.
12
Jacqueline I. Stone, Original Enlightenment and the Transformation of Medieval Japanese Buddhism (Honolulu:
University of Hawai‘i Press, 1999), 125. A regular form of textual consumption, especially in the Tendai school of
Buddhism, was and still is to some extent the kanjin-style interpretations of sacred writings.
13
Ryūichi Abé, The Weaving of Mantra Kūkai and the Construction of Esoteric Buddhist Discourse (New York:
Columbia University Press, 1999), 276.

6
Review of Jeweled-Stūpa Mandala Literature

One of the first studies of the jeweled-stūpa mandalas was conducted by Ishida Mosaku 石田茂

作 in 1941.14 In this pioneering investigation, Ishida provides a brief but solid introduction to the

mandalas of Chūsonji. His account proceeds scroll by scroll, describing the iconography of the

narrative vignettes and focusing on the connection of the scenes to their potential scriptural basis

in the Golden Light Sūtra. A short study by Kameda Tsutomu 亀田孜 follows that of Ishida.15

Kameda looks broadly at the narrative vignettes of the Chūsonji mandalas to draw out

similarities between scenes across the set. He also highlights key narratives and repetitive deities

and shows connections between them and the Ōshū Fujiwara 奥州藤原 (three generations of

rulers likely of Emishi ancestry who, for the most part, independently governed northern

Honshū). Hamada Takashi 浜田隆 has also written on the Chūsonji mandalas. 16 He proposes that

the textual stūpa originates with the popularity of the Lotus Sūtra‘s instructions to build stūpas

for great merit and the Golden Light Sūtra‘s conflation of the stūpa with the Buddha. Much like

the other scholars, Hamada notes stylistic consistencies between narrative vignettes within the

set, such as the treatment of the figures and landscape. He conducts a brief visual comparison

with the Danzan Shrine set and attempts to date the Chūsonji mandalas stylistically and by

historical context.

By far the most extensive examination of the mandalas to date has been conducted by

Miya Tsugio 宮次男 in the tome, Kinji hōtō mandara 金字宝塔曼荼羅 (Golden Script Jeweled-

14
Ishida Mosaku 石田茂作, ―Kokuhō saishōōkyō kyōtō mandara 国宝最勝王経経塔曼荼羅,‖ in Chūsonji
okagami 中尊寺大鏡, vol. 2 (Tokyo: Ōtsuka Kōgeisha, 1941), 4-13.
15
Kameda Tsutomu 亀田孜, ―Jūbun saishōōkyō jūkai hōtō mandara 重文最勝王経十界宝塔曼荼羅,‖ in Chūsonji,
ed. Ishida Mosaku 石田茂作 (Tokyo: Asahi Shinbunsha, 1959), 68.
16
Hamada Takashi 浜田隆, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu 金光明最勝王経金字宝塔曼荼羅図,‖
in Chūsonji 中尊寺, ed. Fujishima Gaijirō 藤島亥治郎 (Tokyo: Kawade Shobō Shinsha, 1971), 145-52, 261-65.

7
Stūpa Mandalas).17 He begins with a contextual introduction to the mandalas in general.

Opening with a discussion of sūtra transcription and stūpa erection practices that relies heavily

on the work of Ishida Mosaku,18 Miya introduces the precedence of sūtra and stūpa combinations

in visual culture. Importantly, he is the first to examine possible prototypes in China and Korea,

pointing out that while the paintings experienced a considerable transformation in Japan, the

textual stūpa first appeared in China and later in Korea. Miya then precedes with much the same

goal as the previous scholars: a close visual analysis of the narrative vignettes surrounding the

central stūpa.

Miya first examines with jeweled-stūpa mandalas of Chūsonji. After providing details of

the mandalas‘ construction, he analyzes the contents of the each scroll‘s narrative vignettes in

turn. Miya provides extensive analysis of the encircling scenes, taking care to identify the deities

and to connect the scenes with the scripture. He then discusses the nature of the narrative

vignettes, arguing that rather than an explanatory style like etoki 絵解 (pictorial decipherment),19

in which images are largely narrative so as to aid the telling of a story known as setsuwa zu 説話

図 (narrativized pictures), where the picture is narrative in quality, the illustrations of the

Chūsonji mandalas are more iconographic and symbolic. He identifies reoccurring deities such

as the Four Guardian Kings (四天王 Jpn. shitennō, Ch. sitianwang; Skt. catur mahā

17
Miya Tsugio 宮次男, Kinji hōtō mandara 金字宝塔曼荼羅 (Tokyo: Yoshikawa Kōbunkan, 1976). Before the
publication of his book, Miya wrote a few articles introducing his ideas which were later incorporated in the
monograph, so these precursory articles are not discussed here.
18
Ishida Mosaku, ―Gangōji gokurakubō hakken no kokerakyō 元興寺極楽坊発見の杮経,‖ in Gangōji gokurakubō
chūsei shomin shinkō shiryō no kenkyū 元興寺極楽坊: 中世庶民信仰資料の研究, ed. Gorai Shigeru 五来重
(Kyoto: Hōzōkan, 1964), 229-38.
19
This is a definition by Ikumi Kaminshi. See Ikumi Kaminishi, Explaining Pictures: Buddhist Propaganda and
Etoki Storytelling in Japan (Honolulu: University of Hawai'i Press, 2006), 3.

8
rājakāyikāḥ)20 and notes that in addition to being important deities in the Golden Light Sūtra,

they also held particular importance in the area controlled by the Ōshū Fujiwara. But precious

little evidence exists to indicate the patron and circumstances of the mandalas‘ production. In

order to formulate a hypothesis about the dating and commission of the mandalas, Miya conducts

a stylistic analysis of the paintings, comparing them to the general characteristics of Heian period

painting and also to other works produced by the Ōshū Fujiwara.

Miya similarly examines the Danzan Shrine set of jeweled-stūpa mandalas. Beginning

with a discussion of a category of paintings known as Lotus Mandalas (法華曼荼羅 hokke

mandara), he notes that the vignettes of the Danzan Shrine set largely conform to this category

of mandalas, which stress the narrative function of pictures—also called transformation tableaux

(変相図 Jpn. hensō zu, Ch. bianxiang tu). He compares the illustrations of the jeweled-stūpa

mandalas with those of celebrated Lotus Mandalas. Miya explains the construction of the

jeweled-stūpa mandalas, noting the density of the illustrations, before dissecting the narratives in

each mandala according to their relationship to Lotus Sūtra chapters and recording the use of

sūtra titles and passages as accompanying cartouches to the scenes. As a supplement to this

involved analysis, Miya provides very clear graphs. Miya also explores the complicated

arrangement of the vignettes and proposes general rules for the placement of the illustrations

across the set. As the cartouches accompanying the pictures are quite extensive, he compares the

style of those of the mandalas with other contemporary examples. Continuing such comparison,

Miya discusses the style and techniques of the Danzan Shrine paintings which he then combines

20
The Four Guardian Kings are as follows: Tamonten of the north (多門天 Ch. Duowentian; Skt. Vaiśravaṇa);
Zōchōten of the south (増長天 Ch. Zengchangtian; Skt. Virūḍhaka); Jikokuten of the east (持国天 Ch. Chiguotian;
Skt. Dhṛtarāṣṭra); and Kōmokuten of the west (広目天 Ch. Guangmutian; Skt.Virūpakāṣya).

9
with the documentary evidence of the mandalas‘ connection with the associated temple Shigaiji

紫盖寺 in Nara to reveal a possible date of production.

Miya‘s final analysis of jeweled-stūpa mandalas concerns those of Ryūhonji. He begins

with an examination of the inscriptions on the backs of the mandalas which indicate the

movements of the mandala set from Hōryūji 法隆寺 in Nara to Ryūhonji in Kyoto during the

Edo period (1600-1868). After describing the construction and general arrangement of the

mandalas, including the configuration of the sūtra text into the stūpa format, Miya explores the

themes and contents of the narratives in much the same manner as his examination of the Danzan

Shrine set. He locates the scenes and passages on each mandala by Lotus Sūtra chapter, again

with explanatory graphs superimposed on the mandalas. Miya concludes by suggesting a date of

production of the thirteenth century based on a stylistic analysis of the mandalas as well as using

information on the painting‘s restoration.

Miya concludes by exploring the three mandala sets as a related collection. He compares

the mandalas stylistically, by the choice of arrangement and narrative vignette selection, and by

the treatment of the cartouches, concluding that the Danzan Shrine set retains more of the older

style of narrative production reminiscent of China while both the Ryūhonji and Chūsonji sets

reflect a move toward a more Japanese-styled interpretation of narrative vignettes. This

engenders a discussion of Danzan Shrine and Ryūhonji sets as Lotus Sūtra paintings, which are

then compared with other such pictures in order to champion the Danzan Shrine and Ryūhonji

sets as critical to the understanding of Lotus Sūtra depiction in Japanese visual culture. Without

Miya‘s solid and thorough research, my project could not have ventured beyond an explanation

of the incredibly complicated visual components of the jeweled-stūpa mandalas. Instead, his

10
work allows me to offer ideas about the possible implications of the imbricated textual stūpa, a

subject largely elided by scholarship on the mandalas to date, including Miya‘s.

Sadly, scholarship after Miya has remained utterly skewed toward the Chūsonji mandalas,

to the neglect of the Danzan Shrine and Ryūhonji sets. Ariga Yoshitaka 有賀祥隆 focuses on

the Chūsonji mandalas‘ unusual combination of blue paper with color illustrations, 21 discovering

that previously only purple paper sūtra copies used color illustrations in the frontispieces, while

the frontispieces of the ubiquitous blue and gold sūtra copies used gold and silver for the

description of the images. The work of Hayashi On focuses primarily on Chūsonji‘s narrative

vignettes, concentrating on reoccurring deity groups and relating them to the faith of the Ōshū

Fujiwara and their importance in the Golden Light Sūtra.22 Aside from the three sets of jeweled-

stūpa mandalas, the discovery of additional, single mandalas separated from their original sets

prompted two more articles. The first article is again by Miya, 23 who analyzes the twelfth-

century jeweled-stūpa mandala at Myōhōji 妙法寺 in Sakai according to his usual method. He

conducts a visual analysis of the narrative vignettes, carefully describing each scene fully and

connecting it to the appropriate chapter of the Lotus Sūtra. Miya then compares and contrasts the

lone mandala with those from the Danzan Shrine and Ryūhonji sets in order to locate the

Myōhōji mandala within the history of the jeweled-stūpa mandalas‘ production. Izumi Takeo 泉

武夫 examines the narrative vignettes of a badly damaged jeweled-stūpa mandala housed in a

private collection and proposes a date for the single mandala in comparison to other examples. 24

21
Ariga Yoshitaka 有賀祥隆, ―Konkōmyō saishōō kyō zu saikō 金光明最勝王経金字宝塔曼荼羅図再考,‖
Chūsonji bukkyō bunka kenkyūjo ronshū 中尊寺仏教文化研究所論集 1 (1997): 92-99.
22
Hayashi On 林温, ―Daichōjuinzō konkōmyōsaishōōkyō kinji hōtō mandara zu oboegaki 大長寿院蔵金光明最勝
王経金字宝塔曼荼羅図覚え書き,‖ Bukkyō geijutsu 仏教芸術 277 (2004): 81-95.
23
Miya Tsugio, ―Myōhōjizō myōhōrengekyō kinji hōtō mandara ni tsuite 妙法寺蔵妙法蓮華経金字宝塔曼陀羅に
ついて,‖ Bijutsu kenkyū 美術研究 337 (1987): 88-96.
24
Izumi Takeo 泉武夫, ―Hokekyō hōtō mandara 法華経宝塔曼荼羅,‖ Kokka 国華 1169 (1993): 29-38.

11
He offers a dating scheme different from Miya‘s and based on an analysis of the density of the

vignettes, concluding that this mandala is actually the earliest known example of Japanese

jeweled-stūpa mandalas, dating to the early twelfth century.

While quite strong, the scholarship in English on the jeweled-stūpa mandalas is even

sparser: only Willa Tanabe and Mimi Yiengpruksawan have discussed the mandalas in any real

detail. In her book, Paintings of the Lotus Sutra, Tanabe analyzes the Danzan Shrine and

Ryūhonji mandalas as examples of the trend, seen at the close of the Heian and beginning of the

Kamakura period, toward an emphasis on narrative description of sūtra content rather than

through text in the art of the Lotus Sūtra.25 Her focus is on the shift in emphasis away from sūtra

text to a reliance on the illustrations to explain the scripture in the jeweled-stūpa mandalas,

seeing them as transitional works linking the conventional blue and gold sūtra illustrated copies

to the completely pictorial transformation tableaux. Most important to Tanabe‘s overall argument

is the miniaturization of the sūtra text and the prominence and primary role of the vignettes.

Mimi Yiengpruksawan examines the Chūsonji jeweled-stūpa mandalas in Hiraizumi: Buddhist

Art and Regional Politics in Twelfth-Century Japan.26 Yiengpruksawan offers a contextualized

study of the mandalas, interweaving the importance of the Golden Light Sūtra to the

authoritative aims of the Ōshū Fujiwara and the intimate illustrations of the narrative vignettes

that reveal the anxieties of the ruling family. She also makes a strong argument for Fujiwara

Hidehira as the patron of the Chūsonji set.

The majority of scholarship concerning the jeweled-stūpa mandalas has been concerned

with formal analysis and iconographic studies of the narrative vignettes. In this regard, the

mandalas have been successfully and thoroughly explicated. Unfortunately, the possible

25
Willa J. Tanabe, Paintings of the Lotus Sutra (New York, Tokyo: Weatherhill, 1988), 98-108.
26
Mimi Hall Yiengpruksawan, Hiraizumi: Buddhist Art and Regional Politics in Twelfth-Century Japan
(Cambridge: Harvard University Asia Center, 1998), 161-84.

12
meanings of the textual reliquary have been largely neglected. Therefore, rather than concentrate

on the narrative vignettes, in this dissertation I focus on the inventive manifestation of the textual

reliquary by questioning the imbrication of sūtra text and architectural stūpa as well as the role

reversal of text and image expressed in the mandalas. In this way, I seek the meanings of the

textual stūpa within the early medieval Buddhist context that reveal the conflation of text, relic,

dharma, and body both in visual culture and in religious praxis.

Plan of the Dissertation

Because the jeweled-stūpa mandalas are a complex web, both visually and conceptually, the

ideas presented in the chapters inevitably overlap and recur. In this way, the dissertation reflects

the visual imbrication of the mandalas, which in turn mimics the conceptual conflation of the

theories underpinning the paintings. One can no more separate sūtra, stūpa, relic, and Buddha

body into discrete parts in the mandalas than in religious practice and doctrine. But in order to

unravel a web, some strand must be pulled first, so I begin in chapter two by situating the origins

of the Japanese jeweled-stūpa mandalas with an examination of the continental prototypes,

followed by an exploration of the culture of sūtra transcription with particular emphasis on the

innovative and intensive copying practices of the eleventh through thirteenth centuries. In this

way, the two investigations locate the continental origin and history of this unusual style of

transcription as well as provide a contextual study of the important trends in sūtra copying

around the time of the mandalas‘ first production in Japan, revealing that, while apparently

singular, the jeweled-stūpa mandalas are nevertheless intimately connected with the larger

movements in the systems of sūtra transcription of the eleventh through thirteenth centuries.

13
 Chapter three explores the formal qualities and commission contexts of the Chūsonji,

Danzan Shrine, and Ryūhonji sets as well as the two lone mandalas. This chapter examines the

challenging construction process involving in the making of these elaborate paintings. I then

analyze the visual qualities of the textual stūpa and the surrounding narrative vignettes, offering

approximate production dates based on examinations of the techniques, styles, and commission

context of the painting sets. On the one hand, this process reveals the singularity of this rare kind

of transcription, but on the other, it underscores their indebtedness to continental models and

Japanese blue and gold illuminated sūtras. The personal contexts of the commissions available

for the Chūsonji set also reveals an intimate portrait of the Ōshū Fujiwara‘s ambitions and fears

manifested through the production of the jeweled-stūpa mandalas.

As sūtra transcriptions—though the scope and magnificence of the mandalas can obscure

the fact that they are just that—the mandalas are based on the belief in the potency of sacred

word. Various and manifold means of accessing the power within dharma were sought at this

time. Chapter four thus explores the professed power of sūtra as proclaimed by scriptures

themselves and in a variety of contemporary records. This investigation recognizes sūtra text as

dharma relic and thus possessing great salvific and apotropaic power. Aside from assertions in

textual records, I look to evidence of this belief in religious practices, such as the burial of sūtras

and the reverent copying of sacred text. I then analyze the combinatory practice of merging sūtra

and stūpa as evidence of text‘s status as dharma relics but also as likely precedents for the

imbrication we see in the mandalas.

Working as tandem chapters, chapter four sets the stage for chapter five‘s discussion of

the salvific matrix of text and body embodied in the mandalas. This chapter continues the

discussion of the mandalas‘ reflection of doctrine and praxis by addressing the question of the

14
stūpa form and denuding it as inextricably connected to Buddha body theory. In the mandalas we

find an imbrication of the bodies of the Buddha, visualized dharmically and architecturally. It is

from this conflation of Buddha body as reliquary and dharma relic that a body of word is

presented. Thus, I look at Buddha body doctrine as the main unifying theory underpinning the

jeweled-stūpa mandalas‘ construction as the visual locus of the text, dharma, body, relic, and

stūpa matrix.

In chapter six, I build on this rare intersection by analyzing the relationships and

functions of word and picture in the jeweled-stūpa mandalas through a semiotic perspective

largely grounded in Buddhism, but aided by Western semiotic theory. While this chapter offers

the first explicit discussion of semiotics, the entire dissertation is based in semiotic concerns over

the nature of representation and the function of text. I begin with an exploration of the

textualized community out of which the mandalas developed, and accordingly discuss a range of

images exhibiting innovative manipulations of text and image in order to demonstrate that the

mandalas depart from the other known text and image relationships of the time. This chapter then

examines the role reversal of text and image and the subsequent issues that arise when viewing

the jeweled-stūpa mandalas. Only upon closer inspection does the viewer become aware that

what once registered as a standard architectural graphic image is actually an elaborate and

precisely choreographed structure crafted of diminutive sūtra characters, one upon the other,

fleshing out the full body of the reliquary. Unlike the legible and tidily spaced characters of

conventionally illustrated sūtras, the text of the jeweled-stūpa mandalas is not meant to be read;

instead, text is experienced, visualizing salvific grace and apotropaic power. Jettisoning its

hermeneutical functions, text functions as image while image is consequently revealed to be text.

The narrative vignettes encircling the textual stūpa also evince a role reversal. With the text no

15
longer readable, these illustrations communicate the signified meaning of the sūtra. Viewers of

the mandalas must read the vignettes in order to read the sūtra and interpret its stories and

didactic lessons. Thus text constructs image in the center of the mandala, while around it image

embodies textuality. Due to the complicated text and image issues at work, this chapter also

investigates the ways in which the audience approaches and views the paintings. One must

negotiate the syncopated viewing experience of the mandalas: registering text as image and

reading image as text.

Conclusion

Whether as artistic complement or graphic usurper, the infusion of dharma relics into the painted

realm represents an important but understudied component of the Buddhist visual lexicon.

Indeed text and image have enjoyed a powerful, complex, and sustained artistic relationship in

Japanese Buddhist painting. Text valued beyond the hermeneutic and the signified comes to

signify itself, crafting an image of salvific power and manifesting the inherent potency and

soteriological sway of both dharma relics and the body of the Buddha in the jeweled-stūpa

mandalas. The complex relationship between word and picture opens new possibilities in the

area of semiotic analysis of Japanese Buddhist painting. Exploring the power of sacred word and

its visual manifestation in the mandalas draws into focus a more comprehensive view of

Buddhist belief, practice, and visual culture during the eleventh through thirteenth centuries.

Sacred word holds an important place in Buddhist visual culture and practice, and scholarly

reflection on this phenomenon can open new lines of inquiry in the field of art history, which is

becoming increasingly interdisciplinary.

16
 Chapter Two

Continental Prototypes and Contexts of Copying

Introduction

This chapter is the first of two which explore the practical matters of the jeweled-stūpa mandalas

such as issues of stylistic origins, the culture of copying at the time of the mandalas‘ production,

the individual histories of the mandala sets, and their formal qualities before venturing on to the

theoretical implications of the imbricated central icon in the remaining three chapters. This

chapter draws on the continental origins as well as the Japanese circumstances that produced the

rare jeweled-stūpa mandalas to reveal that rather than paintings that emerged sui generis for a

brief time in the twelfth and thirteenth centuries, a closer look at the continent and contemporary

Japanese copying practices reveals the mandalas as situated in a system of sūtra copying with

some precedence. Therefore, the chapter begins with an examination of the continental

prototypes, followed by an exploration of the culture of sūtra transcription of the eleventh

through thirteenth centuries, which exposes the trends toward innovative and intensive copying

practices. In this way, I locate the continental source of this unusual style of transcription as well

as provide a contextual study of significant trends in sūtra copying around the time of the

mandalas‘ first production in Japan, revealing that, although highly original, the jeweled-stūpa

mandalas are nevertheless intimately associated with the broader system of eleventh- through

thirteenth-centuries‘ sūtra transcription.

17
Continental Prototypes

Though novel at the time of their first production in twelfth- and thirteenth-century Japan, a

proto-version of the jeweled-stūpa mandalas existed in China as early as the tenth century. The

British Museum houses three such examples, each a small, black inked stūpa built of characters

from the concise Heart Sūtra.1 A closer look at the oldest example 2 from the tenth century

reveals the complex pattern of character arrangement building the wenzi ta 文字塔 or textual

stūpa. The title of the sūtra crowns the stūpa like a canopy: the floating, center line begins with

the characters foshuo 佛説 (sermon of the Buddha), while the rest of the title is split into two

lines. The dangling line to the right of the stūpa continues with bore boluo 般若波羅 and the left

side line concludes the title with miduo xinjing 蜜多心経, together forming Bore boluomiduo

xinjing 般若波羅蜜多心経 (Heart Sūtra). The sūtra begins its seemingly erratic and meandering

course with the first character of the scripture, guan 觀 (meditative insight), located to the center

right of the top line of the foundation. From there the sūtra continues in a straight, diagonal line

down to the left-most character, shen 深 (profound), on the bottom foundational line. Zigzags,

abrupt directional switches, and paths that crisscross over themselves construct the rest of the

visual puzzle. 3 Tracing the outwardly haphazard assembly of sacred characters reveals a complex

pattern of diamonds and triangles. Connecting the dots as it were, even with the assistance of

faint red lines occasionally exposing the trail, is not an easy task. An intimate knowledge of the

1
Jpn. Hannya haramita shingyō; Ch. Bore boluomiduo xinjing; Skt. Prajñāpāramitā hṛdaya sūtra; 般若波羅蜜多
心経; T. no. 251, 8: 848c5-23.
2
For an image of the oldest example, see Miya, Kinji hōtō mandara, 4. The description of the image from Lionel
Giles reads, ―—佛説 [sermon of the Buddha; Ch. foshuo] prefixed to title. Written in a fanciful shape, and with
dotted red lines joining the characters so as to present the outline of a pagoda. Mounted as a kakemono scroll. 22 cm
x 1½ ft. S.5410.‖ See Lionel Giles, Descriptive Catalogue of the Chinese Manuscripts from Tunhuang in the British
Museum (London: Trustees of the British Museum, 1957), 35 entry 1470.
3
Amy McNair has proposed the fascinating idea that the changing directions of the characters might suggest
circulation which in turn recalls the 84,000 atoms of the body. Personal correspondence, September 8, 2011.

18
scripture would be necessary, and given the brevity of the Heart Sūtra, complete memorization

would have been common. But even with the scripture internalized, the path is elusive. Indeed, it

is not until well past the halfway point of the sūtra that the appearance of a random collage of

characters arranged without meaning or order is broken and the interior order, once assumed

structure-less, is revealed to be a patterned system of semantically connected lines of text

symmetrical along the vertical axis. 4 This process would thus require that the puzzle be carefully

devised beforehand.

Miya Tsugio characterizes the complex order as that of a crossword puzzle, the difficulty

of which reminds him of the challenge faced by the Nara-period scholar, Kibi no Makibi 吉備真

備 (695-775).5 According to the legend recorded in the Kibi nyutō setsuwa 吉備入唐説話, a

Tang official gives Kibi no Makibi the poem, Yabataishi 野馬台詩.6 The impenetrability of the

poem required prayers to the deities, Sumiyoshi Myōjin 住吉明神 and Hasedera Kannon 長谷寺

観音, in order to solve the riddle.7 By examining a partially finished textual stūpa in which only

the top portion of the reliquary is drawn with the remainder of the body yet to be written, Miya

concludes that rather than transcribe the sūtra in order, the copyist began at the top of the stūpa

and worked his way down to the foundation. 8 This suggests that the emphasis is not in fact on the

act of copying as religious practice but instead on drawing with text the accurate form of the

stūpa for the creation of a visual puzzle. Thus the emphasis is on deciphering rather than

4
Unfortunately, I have not gained access to the other early examples and so am unable to compare the patterns made
by the accurate connection of the characters. It would be a point of interest to know whether a similar arrangement
of text was used or if new patterns were affected and thus creating new visual games. Giles includes a description of
another example in the British Museum: ―Pan jo po lo mi to hsin ching. Written with dotted lines connecting the
characters so as to form an image of Avalokiteśvara. Verso: Begin. of the same as r° [recto]. Fairly good MS
[manuscript]. Mounted on a scroll. 47 cm x 22 cm. S.4289.‖ See Giles, Descriptive Catalogue, 35.
5
Miya, Kinji hōtō mandara, 5.
6
Ibid.
7
Ibid.
8
Ibid.

19
conventional sūtra transcription. The very nature of the game also precludes reading the text in

any exegetical way. However, the composition of the Heart Sūtra textual stūpa constitutes not

only a fun visual quiz, but also a meritorious action because of the contact with sacred text.

Perhaps deciphering the sacred message acted itself as an amusing memory test since, as several

scholars have shown, strong emphasis was placed on memorizing and internalizing sūtra text in

medieval Japan.9 Moreover, the choice of the stūpa for the textual icon suggests a consideration

of doctrine and praxis, a topic discussed in detail in chapters four and five.

As the earliest example of the textual stūpa format, this tenth-century manifestation is

markedly different from the Japanese versions analyzed in this project. In terms of the character

configuration, whereas the text of the Japanese jeweled-stūpa mandalas continues in an easily

observable order as it constructs the stūpa, the order of the characters in the proto-versions is

intentionally complicated. The puzzle-solving aspect of the textual stūpa was thus largely

abandoned before arriving in Japan. Additionally, while the tenth-century textual stūpa did

require careful pre-planning before its execution, it is hardly on the scale of the elaborate sets

commissioned in medieval Japan. The Japanese mandalas transcribe long sūtras resulting in sets

composing eight to ten large scrolls. The tenth-century Chinese versions are made of less

expensive materials such as paper and black ink, while the Japanese mandalas use costly

resources like large and numerous sheets of dyed blue paper and inks of gold and silver. In light

of these fundamental differences, I do not believe that the earliest examples of the textual stūpa

format were the direct model for the later Japanese mandalas. The textual stūpa developed

further on the continent—and likely in Korea, though no early examples remain—before arriving

9
For a strong example, see Charlotte Eubanks, ―Rendering the Body Buddhist: Sermonizing in Medieval Japan‖
(PhD diss., University of Colorado, Boulder, 2005).

20
in Japan, where the idea was greatly transformed into expensive and involved icons of elaborate

visual beauty and pious intent.

Later examples reveal the development of the textual stūpa format. A tantalizing entry

from the Calligraphy Catalogue of the Xuanhe Period (1119-1125) (Xuanhe shupu 宣和書譜) of

the Northern Song dynasty (960-1127) seems to describe a three-dimensional textual stūpa

illuminated with light.10 According to the entry dated 1112, the Buddhist monk, Fahui 法暉,

presented a spectacular stūpa with sūtra transcriptions in tiny regular script, termed a 細書經塔

xishu jingta, to Emperor Huizong 徽宗 (1082–1135) on the occasion of his birthday as a wish for

longevity. In fact, Fahui was able to accommodate not one but ten scriptures on the stūpa: 11 the

Lotus Sūtra, Śūraṃgama Sūtra,12 Vimalakīrti Sūtra,13 Sūtra of Perfect Enlightenment,14 Diamond

Sūtra,15 Sūtra of Meditation on the Bodhisattva Universal Worthy, Mahākaruṇā puṇḍarīka

Sūtra,16 Dhāraṇī of the Jubilant Corona,17 Dhāraṇī of the Superb Door to an Extended

Lifespan,18 and Sūtra for Humane Kings.19 Placing an incense burner inside the textual stūpa

animated the sacred characters, causing them to fly about before what must have been a

transfixed audience. His piety is credited as the source of his remarkable abilities. Unfortunately,

while this three-dimensional textual stūpa was in the palace collection at the time of the writing

10
Gu Yi 顧逸, ed., Xuanhe shupu 宣和書譜 (Shanghai: Shanghai Shuhua Chubanshe, 1984), 52. Gu Yi punctuated
and collated this text. I am indebted to Professor Amy McNair for this fascinating source.
11
It is possible to translate this passage as ten mandalas illuminated by light from the center.
12
Jpn. Ryōgonkyō; Ch. Lengyanjing; 楞嚴経; T. no. 945, 19: 106b4-155b4.
13
Jpn. Yuimakyō; Ch. Weimo jing; Skt. Vimalakīrti nirdeśa sūtra; 維摩経; T. no. 475, 14: 537a4-557b26.
14
Jpn. Engakukyō; Ch. Yuanjue jing; 圓覺経; T. no. 842, 17: 913a25-922a24.
15
Jpn. Kongō hannya haramitsu kyō; Ch. Jingang bore boluomi jing; Skt. Vajracchedikā prajñāpāramitā sūtra; 金
剛般若波羅蜜経; T. no. 235, 8: 748c18-752c7.
16
Jpn. Daihikyō; Ch. Dabeijing; 大悲経; T. no. 380, 12: 945b4-973a5.
17
Jpn. Butchō sonshō darani kyō; Ch. Foding zunsheng tuoluoni jing; Skt. Uṣṇīṣavijayā-dhāraṇī; 佛頂尊勝陀羅尼
経; T. no. 971, 19: 361c22-364b3.
18
Jpn. Enju myōmon darani kyō; Ch. Yanshou miaomen tuoluoni jing; 延壽妙門陀羅尼経; T. no. 1140, 20: 587c16-
589c22.
19
Jpn. Ninnō gokoku hannya haramitsu kyō; Ch. Renwang huguo banruo boluomi jing; 仁王護國般若波羅蜜経; T.
no. 246, 8: 834a13-845a2.

21
of the Calligraphy Catalogue of the Xuanhe Period, it has not survived and related works neither

exist in material form nor surviving records. But this early twelfth-century example demonstrates

the experimentation with textual stūpas concurrent with and yet vastly different from the

Japanese jeweled-stūpa mandalas.

Later on, the imperial records of the Qianlong Emperor 乾隆帝 (1711-99) of the Qing

dynasty (1644-1911) entitled, Pearl Forest in the Secret Hall,20 document fifty-one textual

stūpas composed from the Song dynasty (960-1279) to the Qing. 21 These crucial records offer

insights into the mysterious production of these rather rare and intricately composed images;

sadly, such a resource is unavailable for the Japanese mandalas studied here. The brief entries

give vital information such as the copyist (including the name when possible), the dynastic date,

the chosen sūtra, and the number of scrolls produced. From this, it is revealed that while not

popularly pursued, persons of elevated rank such as literati and even emperors created textual

stūpas. The most commonly selected sūtras are the Diamond Sūtra with fifteen scrolls and the

Lotus Sūtra with nine scrolls; although the Heart Sūtra is only selected twice, an enthusiastic

Manchu emperor, Shengzu Ren huangdi 聖祖仁皇帝 (1654-1722), also known as the Kangxi

Emperor 康熙帝, configured the scripture into a textual stūpa fifteen times. Other scriptures used

are the Amitābha Sūtra22 with three scrolls, the Scripture of the Original Vows of the Medicine

Master Tathāgata of Lapis Light otherwise known as the Medicine Buddha Sūtra23 with three

scrolls, and the Golden Light Sūtra with one scroll. As mentioned above, the entries are brief in

the Pearl Forest in the Secret Hall, providing valuable but scant information. However, as best

20
Guoli gugong bowuyuan 國立故宮博物院, ed. Midian Zhulin 秘殿珠林 (Taibei: Guoli gugong bowuyuan, 1971).
21
For a list of the fifty-one images complied from the multiple volumes, see Miya, Kinji hōtō mandara, 8-9, n16.
22
Jpn. Amidakyō; Ch. Amituo jing; 阿弥陀経; T. no. 366, 12: 346b26-348b18.
23
Jpn. Yakushi rurikō nyorai hongan kōtoku kyō; Ch. Yaoshi liuliguang rulai benyuan gongde jing; Skt. Bhagavato
bhaiṣajyaguruvaiḍūryaprabhasya pūrvapraṇidhānaviśeṣavistāra; 藥師琉璃光如來本願功德経; T. no. 450, 14:
404c13-408b28.

22
as can be ascertained given the brevity of the passages, seventeen of the textual stūpas recorded

in the Qing text are now housed in Taibei‘s National Palace Museum 國立故宮博物院.24

There still remains the question of connecting these continental proto-versions with the

significantly developed mandalas produced in twelfth- and thirteenth-century Japan. As Miya

points out, two of the fifty-one textual stūpas recorded in the Pearl Forest in the Secret Hall

were prints, and perhaps through printed versions the idea of textual stūpas reached Japan. 25 He

speculates that, at least for the Chūsonji set, a Song dynasty Buddhist print was the likely

model. 26 We might also assume a larger production of textual stūpas than the records confirm

because of the relative ease with which the textual stūpas could be disseminated through prints,

especially when compared to the painstaking transcription by hand-copying of thousands of tiny

characters. It is therefore not hard to imagine that it was through printed copies of textual stūpa

that the concept of textually imbricated stūpas was disseminated to both Korea and Japan. It is,

however, curious that the extant copies and textual records of the Chinese textual stūpas are of a

far simpler variety than the expensively made Korean late Goryeo (918-1392) and Japanese late

Heian and Kamakura examples. And conversely, it is interesting that printed textual stūpas do

not remain from this period. Clearly, broad lacunae characterize the trajectory of the textual

stūpas. But given the scant records concerning the textual stūpas in China and Korea and the

jeweled-stūpa mandalas in Japan, the journey of this uncommon combinatory format is unlikely

to shed its mysterious shroud.

24
For a compiled list of the textual stūpas in the collection of the museum, see Miya, Kinji hōtō mandara, 9 n 17.
Also, National Palace Museum 國立故宮博物院, Gugong shuhuaji 故宮書画録, vol. 8 (Taipei: National Palace
Museum, 1965).
25
Miya Tsugio, ―Kenrantaru kyōten 絢爛たる経典,‖ in Kenrantaru kyōten 絢爛たる経典, ed. Sato Shinji (Tokyo:
Heibonsha, 1983), 96.
26
Miya, Kinji hōtō mandara, 33.

23
 The only Korean example I am aware of is in the collection of Tōji 東寺 in Kyoto.27 And

while the dating and precise provenance of the textual stūpa is uncertain, by calculating the year

mentioned in the vow (願文 gammon) located at the very bottom of the scroll, the date of 1369 is

offered. 28 If so, this places it nearly two centuries beyond the earliest examples of the jeweled-

stūpa mandalas of Japan. However, it seems likely that other examples simply have not survived

or are currently unknown. How precisely the painting came to be in the collection of Tōji is also

unclear. Tōbōki 東宝記, the historical record of Tōji from its founding to the Muromachi period

(1333-1573), documents its existence in the collection by the fourteenth century with a brief

citation recording the existence of an image of a stūpa made from the text of the Lotus Sūtra of

Korean provenance. 29

In contrast to the Japanese versions which portion out the sūtra transcription into the

conventional volume divisions thus making large sets of eight or ten scrolls, this seven-story

Korean stūpa contains the entire Lotus Sūtra.30 And rather than paper, silk dyed a deep blue is

used. Bright, golden characters shine against the blue background. The area enclosing the textual

reliquary is gracefully decorated with bosatsu, flying paradisiacal deities (飛天 Jpn. hiten, Ch.

feitian; Skt. apsarases), worshipers (perhaps portraits of the donors), and flowers that rain down

from heaven, all rendered using fine, gold line. On both sides of each story kneel bodhisattvas

upon lotus pedestals encircled with a thin, golden line and with trailing silver clouds—

surprisingly composed of sūtra characters. In Japanese, these deities are known as kuyō bosatsu

27
For an image with accompanying detail, see Tōji 東寺, ed., Tōji no bijutsu kaiga to kōgei 東寺の美術: 絵画と工
芸 (Kyoto: Tōji, 1976), fig. 34.
28
Miya, Kinji hōtō mandara, 7. Although in a later publication, Miya suggests an earlier date of 1249. See Miya,
―Kenrantaru kyōten,‖ 96-7.
29
Fujita Tsuneyo 藤田経世, ed., ―Tōbōki 東宝記,‖ Kōkan bijutsu shiryō 校刊美術史料, vol. 2 (Tokyo: Chūō
Kōron Bijutsu Shuppan, 1975), 447. I warmly thank Michael Jamentz for bringing this source to my attention.
30
Miya, Kinji hōtō mandara, 6.

24
供養菩薩 or bodhisattvas performing ritual observances. 31 On both sides of the stūpa‘s

foundation and its first floor are drawn two standing figures with halos. Additionally, on the first

floor two identical Buddhas sit side-by-side, their iconography indicative of Śākyamuni and

Prabhūtaratna (多宝 Jpn. Tahō, Ch. Duobao). On each successive story, a Buddha is depicted

emanating rays of light. On each side of the large jewel crowning the finial, paradisiacal deities

fly with outstretched hands of offering. At the bottom of the painting, a vow is written within a

box and flanked by standing, haloed figures; unfortunately, the text of the inscription has

sustained damage over the years, making it difficult to read. But importantly, a passage praising

the combinatory practice marrying sūtra and stūpa is legible; it says that if an image of a stūpa is

made with sūtra text, happiness and great merit will be returned to the practitioner. 32 This rare

direct explanation of the patron‘s ambition in commissioning the textual stūpa illuminates a

fourteenth-century understanding of the vast rewards engendered by the imbrications of sūtra and

stūpa.

Conspicuously absent from the textual stūpas of China and Korea are the narrative

vignettes (経意絵 kyōie) that prominently surround the twelfth- and thirteenth-century Japanese

mandalas. The narrative vignettes seem to be a distinctly Japanese addition but not a consistent

feature after the production of the jeweled-stūpa mandalas of this study, since many later

Japanese textual stūpas lack vignettes. It appears that the Japanese method of textual stūpas

enclosed by sūtra pictorializations was never adopted in China and Korea. Indeed this stark

difference leads Miya to assert that simply referring to the jeweled-stūpa mandalas as mojitō 文

字塔 or textual stūpas is too limiting. Because of the inclusion of graphically-narrativized sūtra

passages rendered in a style similar to that of transformation tableaux, Miya concludes that much
31
Ibid..
32
Ibid., 7.

25
like the broad application of ―mandara‖ to these paintings, the title should also be applied in the

case of the jeweled-stūpa mandalas because to simply classify them as textual stūpas would be to

neglect the pictorializations of the sūtra.33

At this point, it is impossible to know the precise origin or developmental path of the

jeweled-stūpa mandala format. From what can be gathered from the simplified proto-versions

discussed above, the textual stūpa style originally possessed strong indications of a visual puzzle

for the pious and erudite. From the imperial records, it is clear that learned persons, such as

literati and monks, and even emperors copied the scriptures into the form of a stūpa,

demonstrating that this curious style was known and practiced by the educated and elite. But

given that the very process of creating a textual stūpa requires the copyist to be literate,

intimately familiar with the scriptures, and in possession of the texts, the association of the

textual stūpa with the highly ranked levels of society comes as little surprise. This same

connection with the upper echelons continues in the Japanese twelfth- and thirteenth-centuries‘

jeweled-stūpa mandalas, the Chūsonji set being a particularly applicable case. The difference is

that their immense scale and sumptuous artistry necessitates a transfer of brush from elites to

professional copyists and artists.

Culture of Copying

The jeweled-stūpa mandalas, although the product of an elaborate commission requiring great

skill, time, and resources, were nonetheless in both function and intention sūtra transcription

projects. The mandalas served no other ritualistic function, were likely never the main icon (本尊

Jpn. honzon, Ch. benzun) of veneration, and indeed were probably only displayed on rare

33
Ibid.

26
circumstances. However, despite this lack of function beyond the ritual of transcription and the

intention of garnering the consequent merit, the mandalas like many other copying projects were

embedded in a system of meaning where the semiotic expression of sacred word carried its own

contextually specific connotations and the visual combinations of text and image manifested

different Buddhist philosophies. Before discussing the practical aspects, such as the histories of

the mandala sets and formal analyses of the paintings (chapter three), and the theoretical

interpretations of the mandalas (chapters four, five, and six), I discuss here the culture of copying

during the eleventh through thirteenth centuries—a time of burgeoning and tremendous

innovation in sūtra transcription—so as to place the mandalas amongst other inventive projects

in a time that trended toward finding the more extreme and extraordinary forms of sūtra copying.

Prior to this discussion of copying culture in all its forms, the issue of the highly

consistent formatting choice found in the vast majority of sūtra copies needs to be addressed: that

of the seventeen-character line. This character configuration, while ubiquitous in medieval

scripture transcription, nonetheless has a nebulous foundation. Tanaka Kaidō 田中塊堂 explores

sūtra copies‘ conventional arrangement and posits a few explanations; however, the mystery still

largely remains. 34 He begins with a brief examination of symbolic numerology in India, China,

and Japan. Two of his examples offer potential origins for the seventeen-character line. The

Scripture that Transcends the Principle 35claims the number seventeen embodies purity, although

the impact of this declaration is unlikely to dictate such standardization. Alternatively, the odd

number nine is respected as the positive yang (陽 Jpn. myō), while the even number eight is

respected as the inverse yin (陰 Jpn. on) in China. Jointly they total the harmonious seventeen,

the unity of which represents heaven and earth together. Previous scholarship sought answers in

34
Tanaka Kaidō 田中塊堂, Shakyō nyūmon 寫經入門 (Osaka: Sōgensha, 1971), 52-56.
35
Jpn. Hannya rishukyō; Ch. Bore liqu jing; 般若理趣経; T. no. 243, 8: 784a9-786b15.

27
the translation of verses from Sanskrit to Chinese, attempting to reconcile the typical

combination of four, five, or seven characters per verse to the eventual standardization of

seventeen-characters per line. But as Tanaka points out, this theory always leaves unused spaces

when used to formulate a seventeen-character line. 36 Therefore, he returns to the idea of the

numerology associated with principles of yin and yang, positing the seventeen-character line as

most likely indicating the unity of heaven and earth. 37 He finds that by the time of Kumārajīva

the translation of texts into Chinese was standardized in many ways, including that of the

seventeen characters. As this was around the time of the first entrance of scriptures into Japan,

the seventeen-character line was transmitted as well. 38 Tanaka explains the occasional use of

twenty-character lines during the Muromachi period (1333-1573) as an influence from printed

sūtras. However, this highly consistent formatting choice remains relatively constant in most of

the inventive sūtra art discussed below.

Artistic Innovation in Decorated Sūtras (装飾経 sōshokukyō)

A brief introduction to the early history of sūtra copying juxtaposes the differences seen in many

of the projects undertaken during the proliferating complexity of eleventh- through thirteenth-

century sūtra transcriptions. Emperor Shōmu 聖武天皇 (701-56), having established the Office

of Sūtra Reproduction (写経所 shakyōsho), ordered a copy of the complete Buddhist canon (一

切経 Jpn. issaikyō, Ch. yiqiejing) in 734 based on the most updated Chinese Buddhist version

known as the ―Record of Śākyamuni's Teachings Compiled During the Kaiyuan period [712-

756]‖ (開元釋教錄 Jpn. Kaigen shakyō roku, Ch. Kaiyuan shijiao lu), a project yielding more

36
Tanaka, Shakyō nyūmon, 53.
37
Ibid., 53-4.
38
Ibid., 55-6. Tanaka also notes that according to his research, the first time the number seventeen was used in an
official capacity was with the establishment of Shōtoku Taishi‘s seventeen laws. See Ibid., 54.

28
than five-thousand volumes. 39 As the purpose of these early foundational sūtra copies was to

spread the accurate word of Buddhism to the temples across the country, most were written with

black ink on plain paper. 40 These sūtras are known as Tenpyōkyō 天平経, after the date of their

production during the reign of Emperor Shōmu, designated the Tenpyō era (729-49).41

However, while Nara period sūtra transcriptions are not particularly known for their

elaborate decoration, the opulent copies of the Heian period nevertheless had their visual root in

Nara-period examples. For example, when in 741 Shōmu ordered the establishment of the

provincial temple system (国分寺 kokubunji), he mandated that each temple enshrine a copy of

the Golden Light Sūtra in a stūpa.42 Likewise, precious materials were to be used in the copying,

resulting in sūtras written with gold on a purple paper background. So as to accommodate this

immense directive, a special center was opened at the Nara court specializing in gold-lettered

sūtra copies (金字経所 kinjikyō jo).43 These sūtras became known as Kokubunji kyō 国分寺経

and while the sūtra copies in ten volumes were dispersed all around the country, only two

examples remain. 44 Another celebrated Nara-period copy boasting precious materials and an

infamous story is the Flower Garland Sūtra,45 sometimes called the Nigatsudō yakegyō 二月堂

焼経 or the burned sūtras of the Second Month hall. It was on the fourteenth day of the second

39
Shimatani Hiroyuki 島谷弘幸, ―Sōshokukyō no hassei to tenkai 装飾経の発生と展開,‖ in Shin no bi shakyō no
kokoro 信の美: 写経のこころ, ed. Chūsonji (Hiraizumi: Chūsonji, 2000), 18. For a list of transcriptions of the
complete Buddhist canon during the Nara period with dates and patrons, see Ōyama Jinkai 大山仁快, ―Shakyō 写
経,‖ Nihon no bijutsu 日本の美術 156 (1979): 35.
40
Shimatani, ―Sōshokukyō no hassei to tenkai,‖ 18.
41
Tanaka Kaidō, Nihon shakyō sokan 日本寫經綜鑒 (Kyoto: Shibunkaku, 1974), 15.
42
Shimatani, ―Sōshokukyō no hassei to tenkai,‖ 20.
43
Tanaka, Nihon shakyō sokan, 19.
44
Nara National Museum houses Hiroshima‘s Saikokuji‘s 西国寺 set, and Mt. Kōya 高野山 also has one set.
45
Jpn. Daihōkō butsu kegon kyō; Ch. Dafangguang fo huayan jing; Skt. Buddhāvataṃsaka mahāvaipulya sūtra; 大
方広仏華厳経; T. no. 278, 9: 395a4-788b9.

29
month in 166246 that the sūtra got its appellation due to a fire which engulfed the hall at Tōdaiji

東大寺 during the ritual known as shunie 修二会, a repentance ceremony lasting two weeks and

involving both fire and water.47 The scroll was damaged, leaving evidence of the fire along the

bottom of the silver inked sūtra where the blue paper is discolored an orange-brown with shades

of green.

Other evidence of the practice of sūtra copying with decorative paper can be found in the

collection of the Shōsōin 正倉院. Rolls of dyed but unused paper, cut to the size of paper used

for copying, remain as a testament to unfulfilled transcription plans. 48 The variety of colors

among the stored rolls speaks of a creative breadth in sūtra copying during the Nara period.

Textual records also reveal the extent of Nara decorative sūtras. The sixteenth volume of Shōsōin

Documents (正倉院文書 Shōsōin monjo) records sumptuously crafted sūtra papers like purple

paper with gold dust and red paper with silver dust for such scriptures as the Sūtra for Humane

Kings, Original Vows of the Medicine Master Tathāgata of Lapis, Sūtra of the Explication of the

Underlying Meaning,49 and the Lotus Sūtra.50 Shōsōin Documents also record sūtra paper dyed

green, with gold used for the transcription of the sacred word;51 as well as the practice of blue

paper inked with gold and silver,52 so commonly seen in the Heian period. As for illustrated

decorative sūtras produced during the Nara period, the Illustrated Scripture of Cause and Effect

(過去現在因果経絵巻 Kako genzai ingakyō emaki) based on the biographical text of

46
I have opted for a translation of the Japanese reign names into the Gregorian calendar in this project.
47
Nara National Museum 奈良国立博物館, ed., Narachō shakyō 奈良朝写経 (Tokyo: Tokyo Bijutsu, 1983), 51
and fig. 78.
48
Shimatani, ―Sōshokukyō no hassei to tenkai,‖ 19.
49
Jpn. Gejin mikkyō; Ch. Jie shenmi jing; Skt. Saṃdhinirmocana sūtra; 解深密経; T. no. 676, 16: 688b4-711b22.
50
Shimatani, ―Sōshokukyō no hassei to tenkai,‖ 20.
51
Ibid.
52
Ibid., 21.

30
Śākyamuni‘s life, Sūtra of Past and Present Causes and Effects,53 offers an interesting and yet

not often repeated text and image format where the graphic description of the major events in the

Buddha‘s life runs continuously above the text written below. 54 The Illustrated Scripture of

Cause and Effect handscrolls are also recorded in the Shōsōin documents.55 While the eighth

century cannot compete quantitatively or qualitatively with the explosion of decorated sūtras

seen in the eleventh through thirteenth centuries, the origins of decorated sacred word certainly

reach back at least as far as the sūtra productions of the Nara period.

The decorated sūtras of the late tenth through thirteenth centuries experienced a dramatic

increase not only in quantity but also in the variegated manners of production and visual

formatting. The Tale of Flowering Fortunes (栄花物語 Eiga monogatari), an eleventh-century

epic story centered on the life and career of the powerful regent, Fujiwara Michinaga 藤原道長

(966-1028), describes an elaborate scene of courtly copying. During a particularly melancholic

time in the ninth month of 1021, the ladies-in-waiting of Empress Kenshi 藤原賢子 proposed an

ambitious transcription project: each of the attendants, with the addition of close relatives

bringing the participants to the necessary number of thirty, would produce a sumptuous scroll

dedicated to one chapter of the Lotus Sūtra,56 thus creating a thirty-volume set composed of the

twenty-eight chapters of the Lotus Sūtra along with the opening and closing scriptures, Sūtra of

Innumerable Meanings and Sūtra of Meditation on the Bodhisattva Universal Worthy. The

resulting scrolls were quite extravagant. Some composed the sūtra in gold on a blue background;

others incorporated illustrations either above or below the text, directly beneath the text, or as a

53
Jpn. Kako genzai inga kyō; Ch. Guoqu xian zaiyin guo jing; 過去現在因果経; T. no. 189, 3: 620c13-653b28.
54
For a series of eighth-century examples, including a few from the twelfth through fourteenth centuries, see Nara
National Museum, ed., Bukkyō setsuwa no bijutsu 仏教説話の美術 (Kyoto: Shibunkaku, 1996), figs. 2-14.
55
Ibid., 332.
56
Yamanaka Yutaka 山中裕, trans., Eiga monogatari 栄花物語, vol. 2 (Tokyo: Shōgakkan, 1995-98), 233-34.

31
frontispiece. Most of the scrolls were lavishly decorated with the seven treasures (七宝 shippō;

gold, silver, agate, lapis lazuli, coral, crystal, and pearl), and the sūtra rollers and boxes were

bejeweled.57 Upon learning that a location for the sūtra dedication ceremony (経供養 kyō kuyō)

was sought, Michinaga offered the Amida hall (阿弥陀堂 Amidadō) of his temple, Hōjōji 法成

寺, as a stage for the ritual.58 The ceremony seems also to have been a lavish affair with a lecture

praising the ladies and describing their vast rewards as well as chanters of the sūtra title.59 Such

opulent ceremonies and the elaborate sūtra copies embody the longing for paradise through

beautification and elaboration of ritual space and sacred word.

This scene could be the first such event of a large, organized, ritualistic copying and

decorating of sūtra scrolls, for the author expresses that this is the first time to witness such an

astonishing occurrence.60 This particular style of transcription is known as ipponkyō 一品経

(each richly decorated scroll is dedicated to a single chapter of the sūtra). While the scrolls from

this ceremony are not extant, examples like the twelfth-century Kunōjikyō 久能寺経 and Heike

nōkyō 平家納経, dated 1164, offer tantalizing glimpses of what this extravagant project might

have resembled. The Kunōjikyō scrolls derives its name from Tesshūji 鉄舟寺 located on Kunō

mountain 久能山, which is owner of nineteen of the thirty original pieces. 61 During the Edo

period (1600-1868), some of the scrolls were dispersed among the Gotō Art Museum 五島美術

館 in Tokyo (two scrolls), the Tokyo National Museum 東京国立博物館 (three scrolls), and the

57
Ibid., 236-37.
58
Ibid., 234-36.
59
Ibid., 238-42.
60
Tanaka, Nihon shakyō sokan, 21.
61
Komatsu Shigemi 小松茂美, Heike nōkyō no kenkyū 平家納経の研究, vol. 2 (Tokyo: Kōdansha, 1976), 810.

32
Mutō Kinta 武藤金太 collection in Hyōgō prefecture (four scrolls). 62 Slivered, squared, and

sprinkled gold and silver generously decorate the frontispieces and background of the sacred

transcription. Washes of gold and silver lend the scrolls a hazy softness. Much like the scrolls

produced in the Tale of Flowering Fortunes, the preponderance of precious materials

ornamenting the Kunōjikyō reveals its royal associations, for the scrolls are the product of

Emperor Toba 鳥羽天皇 (1103-56), Empress Dowager Taikenmon‘in 待賢門院 (1101-45),

Empress Bifukumon‘in 美福門院 (1117-60), and other aristocrats, and were dedicated in the

twelfth month of 1141.63 Also lavish are the ipponkyō of the Heike nōkyō. Commissioned in

1164 by Taira Kiyomori 平清盛 (1118-81) for dedication at Itsukushima Shrine 厳島神社 on

Miyajima, this elaborate project boasts thirty-three scrolls transcribing multiple sūtras. 64

Kiyomori, writing the petition scroll with his own brush, enlisted thirty-two members of his

family and important retainers to compose a scroll each, resulting in one of the most celebrated

sūtra transcription projects.65 Packed with opulent decoration, the Heike nōkyō layers gold upon

gold with infusions of silver and bright colors. While these two sets are among the finest of their

kind, numerous other examples of scrolls of vibrant colors paired with precious materials survive,

many also of the ipponkyō technique. 66

Compared with these scriptures, most of the decorative sūtra copies produced were not

quite as sumptuous and elaborate, although they were radiant in their own right. The

conventional design took the form of deep indigo dyed paper with gold and/or silver ink for the

62
For images from each of the collections, see Nara National Museum, ed., Hokekyō shakyō to sōgon 法華経: 写経
と荘厳 (Tokyo: Tokyo Bijutsu, 1988), 48-49 plates 106 イ, 106 ロ, 106 ハ, and 106 二.
63
Komatsu, Heike nōkyō no kenkyū, 810.
64
Egami Yasushi 江上綏, ―Sōshokukyō 装飾経,‖ Nihon no bijutsu 日本の美術 278 (1989): 28-29.
65
Ibid.
66
For example, see the scrolls at Hōgonji 宝厳寺 on Chikubushima 竹生島 (eleventh century), Taisanji 太山寺 in
Hyōgō prefecture (twelfth century), Jikōji 慈光寺 in Saitama prefecture (thirteenth century), and Hasedera 長谷寺
in Nara (thirteenth century) to note a few celebrated sets.

33
transcription of the sūtra, a format known as konshi kinginji kyō 紺紙金銀字経 (blue paper, gold

and silver script sūtra), which was often accompanied by frontispiece paintings (見返絵

mikaeshie). This particular type of decorative transcription gained popularity by the tenth century

and continued undiminished throughout the eleventh and twelfth centuries. The oldest remaining

example is the tenth-century, eight volume Lotus Sūtra set housed at Enryakuji 延暦寺 on Mt.

Hiei.67 The frontispiece compositions are rendered in fine-line gold detail, creating a picture with

minimal negative space, while the scripture follows in narrow lines of silver ink. One of the more

celebrated blue-and-gold projects is surely the early eleventh-century Lotus Sūtra copy also at

Enryakuji. This eight-volume set offers a rare view of an early transcription whose lines of

scripture are composed of alternating gold and silver. Because of the great popularity and high

regard of this format, many examples remain from this time of abundant hand-copied

scriptures.68 Of course, decorative paper was not used exclusively for sūtra transcriptions but

often served as the ground for such productions as ornamental collections of literary tales (物語

monogatari) and poetry (和歌 waka).69

The jeweled-stūpa mandalas are rare in their particular design but not necessarily in their

expression of inventiveness because the time surrounding their production saw great momentum

in innovative sūtra art. As shown above, trends toward the decorative in sūtra transcription had a

firm hold by the tenth century. By the eleventh century, copying saw a burst of innovation in text

and image collaboration and a few examples are discussed here in order to establish the fashions

in copying during the eleventh through thirteenth centuries that reveal the mandalas as an

67
For an image, see Nara National Museum, Hokekyō, 43 plate 50.
68
For other notable examples see the decorative scrolls at Honkōji 本興寺 in Shizuoka prefecture (eleventh- and
twelfth-centuries sets), Kongōbuji 金剛峯寺 in Wakayama prefecture (several sets), Rinnōji 輪王寺 in Tochigi
(dated 1129), Itsukushima Shrine (multiple sets), and Hyakusaiji 百済寺 in Shiga, to name just a few. For images of
these scrolls and many more, see Nara National Museum, Hokekyō, 42-45 plates 47-93.
69
See Kyoto National Museum 京都国立博物館, ed., Kana no bi かなの美 (Tokyo: Ōtsuka Kōgeisha, 1996).

34
iteration of a transcription system trending toward more and more inventive designs and at times

extremely intensive practices. Interestingly, the majority of these innovative scrolls are copies of

the Lotus Sūtra—a testament to the scripture‘s great popularity. Rather than retain the structural

chasm between graphic illustration and scriptural text of conventional sūtra copies, word and

picture begin to mingle, as evident in the Ichiji butsu hokekyō (one character, Buddha Lotus

Sūtra scroll; 一字仏法華経) at Zentsūji 善通寺 in Kagawa prefecture.70 In this scroll, a small

drawing of a Buddha seated upon a lotus pedestal is sketched beside each character of the sūtra,

creating alternating lines of ten characters followed by ten Buddhas. The Buddhas are drawn in

black ink with red robes and a seat of green lotus petals, and each figure‘s face and countenance

are depicted differently. The style of the scriptural characters suggests an eleventh-century

date.71

Scrolls such as the Ichiji hōtō hokekyō (one character, jeweled-stūpa Lotus Sūtra scroll;

一字宝塔法華経) which adorn each textual character with a stūpa demonstrate another

manifestation of the expansion of sūtra art at this time. Several of the scrolls made in this style

modify the conventional blue-and-gold transcription type by retaining the pictorial frontispiece

and color scheme while incorporating an enshrining stūpa for the scriptural characters.

Beautifully preserved, the nine scrolls at Honmanji 本満寺, Kyoto produced in the twelfth

century are an excellent example. Against a deep blue, individual stūpas vividly expressed with

luminescent silver for the body and pedestal and fine gold detail for the finial (相輪 sōrin)

enthrone the sacred characters composed in generous gold. The stūpas of the Ichiji hōtō hokekyō

70
For an image, see Nara National Museum, Hokekyō, 283 plate 118.
71
Egami, ―Sōshokukyō,‖ 37.

35
format range from highly individualized and detailed, like those of the Honmanji scrolls, 72 to the

cursory and abbreviated, like the scrolls dated to 1163 and commissioned by the monk, Shinsai

心西,73 in the Nara National Museum (one scroll) 74 and in the private collections in Tokyo of

Sorimachi Kyōsaku 反町恭作 (two scrolls) and Hattori Shōji 服部正次 (one scroll).75 This

format also employed decorative paper like the twelfth-century scroll of Togakushi Shrine 戸隠

神社 in Nagano prefecture, using light grey paper adorned with mica powdered stūpas

enshrining individual characters of black ink thought to have been written by Fujiwara Sadanobu

藤原定信 (1088-1156) because of the slanted style of calligraphy. 76 And in the typical style of

the Heike nōkyō scrolls at Itsukushima Shrine, the Lotus Sūtra‘s ―Apparition of the Jeweled-

Stūpa‖77 chapter (c. 1164) is composed on ornamented paper embellished with gold and silver

and each character drawn within a stūpa. 78

Another format corresponding to this type of inventive copying is the Ichiji rendai

hokekyō 一字蓮台法華経 (one character, lotus pedestal Lotus Sūtra scroll) in which each

character rests upon a lotus pedestal. The two scrolls of the eleventh or twelfth century in Kyoto

National Museum79 and the nine in the collection of Ryūkōji 竜光寺 in Fukushima, believed to

be from the same original set, depict a complex pattern of coordinated lotus pedestal colors. 80

72
Others of this type include the scroll at Rinnōji (twelfth century). For images of these scrolls, see Nara National
Museum, Hokekyō, 280 plates 114 and 115.
73
Nara National Museum, ed., Seichi Ninpō (Ninpō) Nihon bukkyō 1300-nen no genryū 聖地寧波 (ニンポー): 日
本仏教 1300 年の源流 (Nara: Nara Kokuritsu Hakubutsukan, 2009), 299.
74
For an image of the scroll in the Nara National Museum, see Ibid., 80 plate 75.
75
For images of these scrolls in the private collections, see Nara National Museum, Hokekyō, 277-79 plates 113 イ
and 113 ロ.
76
Egami, ―Sōshokukyō,‖ 40. For an image, see Nara National Museum, Hokekyō, 281 plate 116.
77
Jpn. Ken hōtō bon; Ch. Jian baota pin; 見宝塔品.
78
For an image, see Egami, ―Sōshokukyō,‖ 40 fig. 48.
79
For images of the scrolls in the collection of Kyoto National Museum, see ―Ichiji rendai hokekyō,‖ Kyoto
National Museum, accessed August 7, 2011, http://www.kyohaku.go.jp/jp/syuzou/index.html.
80
For an image of the Ryūkōji scrolls, see Nara National Museum, Hokekyō, 282 plate 117.

36
For instance, the lotus pedestals of chapter twenty-one of the Lotus Sūtra in the collection of the

Kyoto National Museum are arranged in rotating colors along the horizontal lines of text

beginning with pale blue and followed by red, green, and silver 81 moving from left to right.

Chapter twenty-two, also in the Kyoto National Museum, further complicates the color

arrangement producing a pattern of interwoven color in the form of a diamond. 82 The twelfth-

century Ichiji rendai hokekyō in the collection of Nara‘s Yamato Bunkakan 大和文華館

museum is a highly ornamented scroll making use of large amounts of gold and silver and a full-

color frontispiece illustration. 83 The lotus pedestals, colored white, cinnabar, and blue-green,

enthrone each character of the scripture. The handwriting is thought to be that of Go-Shirakawa

後白河天皇 (1127-92), and correspondingly the central aristocratic figure in the frontispiece is

believed to represent the emperor with his consort seated slightly behind him and at an angle in a

scene of gathered monks and aristocrats chanting the Lotus Sūtra.84 The scrolls of the Ichiji butsu

hokekyō, Ichiji hōtō hokekyō, and Ichiji rendai hokekyō all demonstrate an elaboration on

conventional sūtra transcription formats and represent the contemporary trend of seeking

increasingly inventive ways of copying the scriptures.

The Lotus Sūtra fans (扇面法華経冊子 senmen hokekyō sasshi)85 the Lotus Sūtra

booklets (法華経冊子 hokekyō sasshi),86 and the Menashikyō 目無経 (literally, the ―eyeless

81
And as a twist, each line of silver alternates white; see Kyoto National Museum, Koshakyō seinaru moji no sekai
古写経: 聖なる文字の世界 (Kyoto: Kyoto National Museum, 2004), 313-14.
82
Ibid.
83
For an image, see Nara National Museum, Hokekyō, 47 plate 102.
84
Kyoto National Museum, Koshakyō, 315.
85
Lotus Sūtra fans are held in the collections of Tokyo National Museum; Idemitsu Museum of Arts 光美術館 in
Tokyo; Saikyōji 西教寺 in Shiga prefecture; Fujita Museum of Art 藤田美術館 in Osaka; Hōryūji; two private
collections; and the largest amassment in the collection of Shitennōji 四天王寺 in Nara. For images of these fans,
see Nara National Museum, Hokekyō, 261-74 plates 112 イ, 112 ロ, 112 ハ, 112 二, 112 オ, 112 へ, and 112 ト.
86
Examples of this type can be found in the Gotō Museum of Art and in the private collection of Ueno Jun‘ichi 上
野淳一. For images of these booklets, see Ibid., 257-60 plates 110 and 111.

37
sūtra‖),87 all of the twelfth century, reveal an increased interaction between scripture and picture,

embodying the fashion in sūtra art which sought new and elaborate designs. While the formats

take the shape of fans, booklets, and scrolls, the layering of sacred script atop images of the

secular world is a feature consistent throughout all the productions and one utterly novel to the

world of sūtra art at the time. Visible beneath the tidy characters are pictures of a world far less

orderly and in need of the redeeming power of sūtras. As such, they stand as inventive

elaborations upon the conventional design of the transcription of scriptures.

Extreme Practices in Sūtra Transcription

Sūtra transcription practices in the eleventh through thirteenth centuries also demonstrate a

heightened complexity reflecting the general trends in copying at this time. Whether evinced in

terms of sheer quantity, pace, genuflection, interment, or alternative media, the religious practice

of copying became increasingly imaginative and complicated, much like the sūtra art discussed

above. Although I cannot comprehensively survey all forms of intensive copying here, I have

chosen emblematic manifestations of extreme exercises to reveal the parallel between religious

practice and the visual inventiveness seen in art of the time. By doing this, I expose the context

of the jeweled-stūpa mandalas‘ creation as one of novelty in artistic manipulations and religious

practices of transcription, which suggests the mandalas are a manifestation of these phenomena.

87
The remains of the Golden Light Sūtra are in the collections of the Tokyo National Museum which owns a
fragment and the Kyoto National Museum which owns volumes two, three, and four. For the scrolls in Tokyo, see
―Konkōmyōkyō Volume Four Menashikyō‘,‖ Tokyo National Museum, accessed August 7, 2011,
http://www.tnm.jp/modules/r_collection/index.php?controller=dtl&colid=B2400&t=type_s&id=24&lang=ja. For
the scrolls in Kyoto, see ―Konkōmyōkyō,‖ Kyoto National Museum, accessed August 7, 2011,
http://www.kyohaku.go.jp/jp/syuzou/index.html. The Scripture that Transcends the Principle version is in the Dai
Tōkyū Memorial Library 大東急記念文庫 in Tokyo. For images of this version, see Marsha Weidner, ed.,
Flowering in the Shadows: Women in the History of Chinese and Japanese Painting (Honolulu: University of
Hawai‘i Press, 1990), 168-69 figs. 1-3.

38
Quantity

While the practice of copying the entire Buddhist canon dates back to the seventh century, the

exercise increased in popularity and prevalence during the eleventh through thirteenth

centuries. 88 Another notable difference in the Buddhist canon productions at this later time was

that more individual people and small groups of family and friends undertook to copy by hand

such enormous projects, likely upheld by the belief that vast quantity and effort are rewarded by

great merit,89 although aristocrats and imperial family members also continued the commission

of the Buddhist canon, even producing several copies in the expensive blue-gold technique.90

One of the earliest examples of lay individuals engaging in the Buddhist canon

production at this time comes from an 1106 entry in Chūyūki 中右記, the diary of Fujiwara

Munetada 藤原宗忠 (1062-1141). According to the record, an unnamed holy person from Tōji

walked Kyoto encouraging residents to copy the entire Buddhist canon, eventually copying a set

and conducting the dedication service at a hall of Emperor Shirakawa 白河天皇 (1053-1129).91

A similarly vague entry can be found in Hyakurenshō 百錬抄, a thirteenth-century anthology of

various records and tales by an unknown compiler. 92 On the first day of the sixth month in 1115,

another unnamed holy person at Kitano 北野 copied and performed the dedication of a Buddhist

88
Tsuji Zennosuke 辻善之助, Nihon no bukkyō shi 日本佛敎史, vol. 1 (Tokyo: Iwanami Shoten, 1960), 659.
Because Tsuji analyzed and amassed a tremendous quantity of information about Japanese Buddhist religious
practices from a host of primary documents, this publication is of great value for researchers, despite the absence of
broader analysis and dating errors in the translation of Japanese reign dates to the Gregorian calendrical system.
89
Tsuji, Nihon no bukkyō shi, 659.
90
For instance, Fujiwara Kiyohira 藤原清衡 (1056-1128) commissioned a blue paper, gold and silver script copy of
the Buddhist canon in ca. 1117 known as the Kiyohirakyō 清衡経; Emperor Toba 鳥羽天皇 (1103-56)
commissioned in the mid-twelfth century a blue-gold copy of the Buddhist canon now known as the Jingōjikyō 神護
寺経 for Go-Shirakawa; Bifukumon‘in commissioned the set known as the Arakawakyō 荒川経 in 1150 for the
repose of Emperor Toba‘s soul; and Fujiwara Hidehira 藤原秀衡 (1122-1187) completed in ca. 1176 a blue-gold
copy of the Buddhist canon.
91
Tsuji, Nihon no bukkyō shi, 659.
92
Kuroita Katsumi, ed., ―Hyakurenshō 百錬抄,‖ in Shintei zōho, Kokushi taikei, vol. 11 (Tokyo: Kokushi Taikei
Kankōkai, 1929), 50. Tsuji makes the mistake of saying the event occurred in 1117. See Tsuji, Nihon no bukkyō shi,
659.

39
canon. Honchōseki 本朝世紀, a mid-twelfth-century text compiled by Fujiwara Shinzei 藤原信

西 (1106-60), also records that in 1143 the monk, Kaku‘a 覺阿, copied the Buddhist canon. 93

Many other such examples exist, but probably the most renowned instance of the transcription of

the Buddhist canon by an individual is that of Fujiwara Sadanobu.94 Vowing at the age of forty-

two to copy by hand the entire Buddhist canon, Sadanobu finally finished the massive project

twenty-three years later at the age of sixty-four. Tsuji Zennosuke 辻善之助 estimates that the

endeavor thus required Sadanobu to copy around two volumes every three days. 95 So celebrated

and astonishing was this undertaking that it is recorded with great amazement and praise in

multiple medieval texts. For instance, the twelfth-century text of disputable authorship,

Imakagami 今鏡, extolled Sadanobu‘s dedication for copying the Buddhist canon with his own

brush (also known as ippitsukyō 一筆経, or sūtra copied with one brush), remarking that he does

not seem to be an ordinary person and that one never hears of another quite like him. Fujiwara

Yorinaga 藤原頼長 (1120-56) commended Sadanobu in his diary, Taiki 台記, writing that the

enormity of the project will ensure Sadanobu‘s name in history.96 He also effused that in the past,

present, and even the future no one will be able to accomplish an equivalent feat. 97 As a gesture

of his respect for such efforts, Yorinaga donned new robes and washed his mouth before meeting

with Sadanobu.98

Tales remain of others in less financially and well-connected circumstances vowing to

copy the Buddhist canon. The mendicant monk known commonly as Shikijō 色正 enlisted the

93
Ibid., 659.
94
The following information about Sadanobu‘s project is based on Tsuji‘s research unless otherwise noted. See Ibid.,
660-61.
95
Ibid., 660.
96
Komatsu Shigemi, ―Ichiji sanrei no shakyō 一字三礼の写経,‖ Museum 186 (1966): 3.
97
Tanaka, Nihon shakyō sokan, 24.
98
Komatsu, ―Ichiji sanrei no shakyō,‖ 3.

40
aid of his fellow monks, Saikan 西観 and Shinshō 心昭, in begging for paper, brush, and ink

during their travels in order to fulfill the ambitious vow. 99 Having bathed himself in incense,

Shikijō set himself to the task of copying the canon. The project began in 1187 when he was

twenty-nine year old and was not completed until 1228 when Shikijō was 70 years of age, taking

a total of forty-two years. Tsuji again provides calculations for the labor, estimating that in the

span of one month Shikijō copied around ten volumes and so averaged one volume every three

days. Based on the inscriptions, it is possible to see the circumstances under which the diligent

group toiled. Shikijō records that in their journeys all over the country, even while standing,

walking, or on a boat, he copied the sūtras. Of the original 5048 volumes, over four thousand

survive in the collection of Kōshōji 興正寺 in Tajima, Kanagawa, despite 448 which were

spoiled by insects and a severe flood in 1702 that damaged 1200 volumes, 230 of them fatally.

Pace

Another hallmark of the intensification of ritualistic copying was the extreme pace set by some

performances. It was not uncommon for large groups of people to assemble so that they might

collectively copy substantial quantities of scriptures all together in just one day. On the fifth day

of the fifth month in 1135, Emperor Toba commissioned all 600 fascicles of the Perfection of

Wisdom Sūtras100 to be copied in just one day at Hosshōji 法勝寺, Kyoto.101 Not content with

this massive effort, devotees attempted even more astonishing copying feats.

99
The following information about Shikijō‘s project is based on Tsuji‘s research. Tsuji, Nihon no bukkyō shi, 661-
62.
100
Jpn. Dai hannya haramitta kyō; Ch. Da bore boluomiduo jing; Skt. Mahaprajñāpāramitā sūtras; 大般若波羅蜜
多経; T. no. 220, 5: 1a4-7: 1110b3.
101
Tsuji, Nihon no bukkyō shi, 663.

41
 Probably one of the most daunting and logistically challenging types of sūtra

transcription is copying the entire Buddhist canon of over 5000 volumes in a single day, known

as ichinichi issaikyō 一日一切経. But just such an event occurred on the eighteenth day of the

third month in 1096 when ten thousand people from all literate strata of society gathered in

Kyoto to copy the canon. 102 In 1211, on the twenty-third day of the fourth month, an ichinichi

issaikyō event was organized by Emperor Go-Toba 後鳥羽天皇 (1180-1239) at his recently

constructed temple, Saishō Shitennō‘in 最勝四天王院.103 Monks from all around the country,

totaling 13, 215, congregated in Kyoto for the massive service, all under the sponsorship of the

emperor. According to multiple sources, the result was an unparalleled event. 104 These

performances of extreme sūtra transcription practices once again reflect the drive to reach new

heights in copying typical of the eleventh through thirteenth centuries.

Genuflection

The laborious practice of ichiji sanrei 一字三礼 (one character, three bows), in which the

copyist writes one character and then pays obeisance three times, usually understood to be

performed as bows, before moving on to the next character,105 is another manifestation. Two

notable examples of this practice, including the related ichigyō sanrei 一行三礼 (one line, three

bows) in which obeisance is paid to each line of characters copied, were carried out by the

Buddhist sculptor from the Kei school 慶派, Unkei 運慶 (1151–1223), and the courtier,

102
Ibid.
103
Ibid., 664.
104
Ibid.
105
Nakamura Hajime 中村元, Iwanami bukkyō jiten 岩波仏教辞典 (Tokyo: Iwanami Shoten, 1989), 30.

42
Madenokōji Nobufusa 万里小路宣房 (1258-1326).106 In 1183, Unkei made a vow to copy the

Lotus Sūtra according to very strict procedures.107 Fortunately, the inscriptions on the scrolls

illuminate much about the nature of the mission. Elaborate efforts were made to guarantee the

purity of the process. According to the inscription on the eighth volume, participants ensured the

cleanliness of their bodies and clothes; the paper was specially made; the scroll rollers were

crafted from the wood remaining after Taira Shigehira 平重衡 (1158–1185) razed Tōdaiji 東大

寺 in Nara; and water for the ink was drawn from three different sacred places: Miidera 三井寺,

the Yokokawa 横川 on Mt. Hiei 比叡, and Kiyomizudera 清水寺. Fifty men and women,

including another celebrated sculptor from the Kei school, Kaikei 快慶 (late twelfth or early

thirteenth century), participated in the project. And on top of the extraordinary lengths Unkei

took to guarantee the sacredness of the scrolls (also an indication of the overall trends in copying

in that Unkei was thinking not only of ways to intensify the practice of the copying and the

exterior appearance of the scrolls, but also of the interior composition), after each line of text was

copied, three bows were made to the recently finished characters. Unkei tabulated the number of

bows, nenbutsu 念仏 chants (calling on the name of Amida Buddha [阿弥陀 Ch. Amituo; Skt.

Amitābha]), and chanting of the august title of the Lotus Sūtra (daimoku 題目) that the project

required: 50,000 bows, 100,000 nenbutsu chants, and 100,000 chants of the title of the Lotus

Sūtra. And in order to prevent an invasion of demons, every day services were performed and ten

parts of the Lotus Sūtra were read.

106
Tanabe offers a parallel example: Emperor Reizei 冷泉 (950-1011), while copying several sūtras, prayed after
writing each character as recorded in the Honchō monshū 本朝文集. Tanabe, Paintings of the Lotus Sutra, 46. See
Kuroita Katsumi 黑板勝美, ed., ―Honchō monshū 本朝文集,‖ in Shintei zōho, Kokushi taikei 新訂増補, 國史大系,
vol. 30 (Tokyo: Kokushi Taikei Kankōkai, 1929), 205.
107
The following information concerning Unkei‘s project comes from Tsuji, Nihon no bukkyō shi, 672-73. See also
Komatsu, ―Ichiji sanrei no shakyō,‖ 4-8.

43
 A later example is that of Nobufusa, who copied two-hundred volumes of the five great

Mahāyāna sūtras (五部大乘経 Jpn. gobu daijō kyō, Ch. wubu dasheng jing) using the ‗one

character, three bows‘ technique.108 In several of the inscriptions, it becomes clear that Nobufusa

undertook this challenging mission not only to generate merit for himself, but also for his parents.

In the seventh volume of the Lotus Sūtra, he writes that this volume was dedicated as a memorial

to a deceased family member. Komatsu Shigemi 小松茂美 identifies this person to be

Nobufusa‘s father, who retired from public life to join the Buddhist ranks in 1284 due to illness,

but was fortunate enough to live for an additional twenty years. The inscription coordinates with

the seventh anniversary of his father‘s death, and the third volume of the Great Collection

Sūtra109 he dedicated to his deceased mother.110 Such laborious genuflection corresponds to the

search for more inventive and challenging ways of creating sūtra copies.

Alternative Media

The move toward innovation was also reflected in the incorporation of alternative media.

Although there were many other types of media employed, I want to highlight here the cases of

stone sūtras, tile sūtras, and blood copying. The practice of copying sūtra text onto stone is

known as sekkyō 石経. This term refers to the broad practice of copying scripture onto the

durable surface of stone and is more commonly ascribed to the longstanding tradition of copying

sūtras onto stone tablets.111 However, it also includes the more uncommon practice of inscribing

a single character onto each stone, known as isseki ichijikyō 一石一字経, or of inscribing several

108
The following information concerning Nobufusa‘s efforts comes from Ibid., 3-8.
109
Jpn. Daijikkyō; Ch. Dajijing; Skt. Mahāsaṃnipata sūtra; 大集経; T. no. 397, 13: 1a4-407a17.
110
For more examples of this phenomenon, see Komatsu, ―Ichiji sanrei no shakyō,‖ 4.
111
For more on this topic, see Kuno Takeshi 久野健 and Nakamura Hajime, eds., Bukkyō bijutsu jiten 仏教美術事
典 (Tokyo: Tokyo Shoseki, 2002), 499-500.

44
characters per stone, referred to as taji isseki 多字一石.112 The small stones often measure

between three to ten centimeters in diameter. The text of the sūtra is frequently written in black

or red ink. 113 Because of the nature of the small stone transcription, even though the stones might

all be completed and stored together, which often meant burial, the sūtra could not be

reconstructed without a superhuman feat of will and copious amounts of time, thus

reconstruction was never the point.114 Tile sūtras, or kawarakyō 瓦経, present a similar

situation. 115 Typically measuring thirty centimeters, the ceramic tiles are scored with a sharp

implement to carve the lines for the sūtra text—much like the lines of conventional sūtras—and

then the scriptural lines are copied, often on both surfaces of the tile, while the sūtra title and

volume number are inscribed on the sides. After their firing in a kiln, the tile sūtras were often

buried standing up in the ground with a stūpa sometimes marking the site. 116 Occasionally, rather

than sūtra text on both sides of the tile, one side might have rows of Buddha images, resembling

the Ichiji butsu kyō.117 In 1142, the Shingon monk, Zen‘ne 禅恵, began copying sūtras onto tile,

producing five hundred by the following year. 118 Zen‘ne began this project with a rather long list

of vows he hoped to fulfill with the merit generated from the tile sūtras and sculptures: grand

prayers for the nation‘s and emperor‘s peace as well as more intimate appeals for his own peace

in this realm, a long and healthy life of good quality, and to be reborn into paradise. 119 Together

with the Amida and Jizō 地蔵 (Ch. Dizang; Skt. Kṣitigarbha) sculptures he made, the tiles were

112
These terms come from Ikemi Sumitaka 池見澄隆, ―Tsumi to sono kaiketsu 罪とその解決,‖ in Hokekyō no
shinri: sukui o motomete 法華経の真理: 救いをもとめて, ed. Miya Tsugio (Tokyo: Shūeisha, 1989), 129. An
image of the small stone sūtras can also be found on this same page.
113
Ibid.
114
For more on this topic, see Seki Hideo 関秀夫, ―Kyōzuka to sonno ibutsu 経塚とその遺物,‖ Nihon no bijutsu
日本の美術 292 (1990), 70-79.
115
For an image of a title sūtra, see Ibid., 128.
116
Tanaka, Nihon shakyō sokan, 27.
117
Ikemi, ―Tsumi to sono kaiketsu,‖ 128.
118
Ibid., 129.
119
Ibid.

45
buried at his family‘s mountain temple. 120 Tanaka Kaidō explains that the burial of sacred text

purifies the land, and as the land is the source of all including the nation, the purified land and its

inhabitants are united.121

Copying scriptures in blood, while not that common, represents one of the more intimate

and extreme forms of sūtra transcription.122 Fujiwara Yorinaga famously copied sūtras in blood,

although not wanting to use his own, he asked Fujiwara Atsuto to make a sanguinary donation

for the project.123 According to the Tale of the Hōgen Disturbance (保元物語 Hōgen

monogatari), the exiled Emperor Sutoku‘in 崇徳天皇 (1119-1164) wrote scriptures in ink mixed

with his own blood for three years in hopes of securing a paradisiacal birth after death. 124

Practitioners of scriptural blood writing seek to transform what is illusory into something

adamantine, hence blood into dharma. 125 Blood was not the only substance capable of

establishing a karmic bond; Fujiwara Munetada in 1136 enshrined votive copies of sūtras that he

and his children transcribed on paper containing strands of his deceased wife‘s hair. There are

even those tales of the Buddha‘s former lives (闍多伽 Jpn. jataka, Ch. sheduoqie; Skt. jātaka)

that describe the self-flaying of skin for paper, liquefying of marrow and pulverizing of flesh for

ink, and the breaking of bones for brushes, all so that sacred word can be copied.

Not content with mere paper and ink, alternative media such as small stone sūtras, tile

sūtras, and blood copying represent the search for new and inventive means to transcribe

scripture. As with the other examples provided in this section, while the trend encouraged

120
Ibid.
121
Tanaka, Nihon shakyō sokan, 27. For more on the practice of sūtra burials, see chapter three.
122
For more on this practice, see John Kieschnick, ―Blood Writing in Chinese Buddhism,‖ Journal of the
International Association of Buddhist Studies 23 (2000): 177-94.
123
Tanabe, Paintings of the Lotus Sutra, 56.
124
Nagazumi Yasuaki 永積安明 and Shimada Isao 島田勇雄, eds., Hōgen monogatari; Heiji monogatari 保元物
語; 平治物語, in Nihon koten bungaku taikei 日本古典文學大系, vol. 31 (Tokyo: Iwanami Shoten, 1961), 179.
125
Rambelli, Buddhist Materiality, 119.

46
copying in novel and innovative ways, the drive was often to establish more personal

connections with the sūtra and its salvific and restorative power by undergoing extreme measures

and even by merging the materially intimate with the numinous nature of scripture. Certainly,

these are but a few of the examples and possible categories of extreme copying. But in selecting

these samplings of intensified scripture transcription practices, strong parallels can be seen with

the art of sūtra copies, revealing the overall trend toward the extreme that, importantly, also

typifies the production of the jeweled-stūpa mandalas.

Fundamental Functions of Sūtra Transcription

The more explicit reasons for performing sūtras in myriad manifestations are analyzed in chapter

four on dharma relics; however, I introduce here the basic concepts driving the faithful to copy

sūtras, sometimes to elaborate degrees and in extreme circumstances. The examples of sūtra

transcriptions examined above represent a type of copying known as kechienkyō 結縁経 or

sūtras that establish kechien 結縁, a connection between the copyists and patrons with the

Buddha, thus bequeathing great karmic merit for the hope of future salvation. 126 The earliest

mention of the term kechienkyō comes from the diary of the Heian-period courtier, Fujiwara

Sanesuke 藤原実資 (957-1046), known as Shōyūki 小右記127 and occurred in the ninth month

and tenth day of 1021.128 The term occurs with frequency after this point, and another example

merging Buddhist canon copies and kechien ceremonies comes from Hyakurenshō. On the fourth

day of the third month in 1142, a ceremony utilizing a copy of the Buddhist canon was held at

126
Nakamura, Iwanami bukkyō jiten, 226.
127
Fujiwara Sanesuke 藤原実資, Shōyūki 小右記, in Dai nihon kokiroku 大日本古記錄, vol. 6 (Tokyo: Iwanami
Shoten, 1971).
128
Egami, ―Sōshokukyō,‖ 19. This occurs in reference to the transcription and dedication service of the Lotus Sūtra
at Muryōju‘in 無量壽院 sponsored by the Empress Fujiwara Kenshi 藤原妍子 (994-1027). See Fujiwara, Shōyūki,
46.

47
the Byōdōin 平等院 in Uji in order to establish kechien for the benefit of Emperor Toba.129 In

transcription performances reminiscent of the Heike nōkyō and the scene from the Tale of

Flowering Fortunes, the typical arrangement began with a rather large group of people in which

each person prepared a single scroll and concluded with the dedicatory ritual of the sūtras as a

completed set.130 However, if the projects lacked participants, then a person was assigned more

than one scroll. 131 The sūtra dedication ceremony described above imbued recently copied sūtras

with the essence of the Buddha, thereby in a sense activating them and solidifying the connection

between the participants and the Buddha. Fabio Rambelli notes that ―texts were endowed with all

the characteristics of sacred objects and were not essentially different from relics, icons, and

talismans‖132 and that ―[a]s soteriological tools….[t]hey acquired a magical and mystical

dimension as sorts of ‗relics‘ of past masters (and ultimately, of the Buddha).‖ 133 Much as icons

and stūpas doubled for the Buddha in the illusory realm, sūtras were not merely symbols of the

Buddha‘s presence, but rather were embodiments of the Buddha. 134 The same karmic connection

is possible in the more intimate and personal copying rituals described. The ornamentation of

scriptures, the inclusion of bodily material, and the labors of the hand to copy sacred word all

establish personal and lasting connections with the numinocity of the dharma through tactile

transference.

As with the jeweled-stūpa mandalas, the primary function of these sūtra transcriptions is

fulfilled in the act of copying itself. The merit from the reverential treatment of the scriptures and

the karmic connection established through the textual contact and labor exerted is earned in the

129
Kuroita, ―Hyakurenshō,‖ 65.
130
Egami, ―Sōshokukyō,‖ 19.
131
Ibid.
132
Rambelli, Buddhist Materiality, 90.
133
Ibid., 96.
134
For a discussion of embodied Buddhist visual and material culture, see Robert H. Sharf, "On the Allure of
Buddhist Relics," Representations 66 (1999): 75-99.

48
moments of the practice of copying, in that direct connection with the scripture in the case of a

personal, hand-copied sūtra, and in the commission and facilitation of copying in the case of

patrons. This is the case even with projects that clearly exhibit a puzzle-like component to their

transcription, such as the tenth-century Chinese textual stūpa and Heike nōkyō, as will be shown

in the sixth chapter. This is not to negate or diminish the further lives of the sūtra copies, or even

the merit they continued to generate, but to emphasize that the very act of transcription was the

religious goal, although a certain level of social prestige and love of beauty must have factored

into the creation of sūtra art as well. But as in cases such as the jeweled-stūpa mandalas where

the scrolls were stored away and rarely presented in any ritualistic context, the production of the

sūtra transcription itself embodied the fundamental function of the project. However, as explored

in chapters four and five, the mandalas‘ symbolic and theoretical functions extend beyond the

original merit gained and karmic connection established in the moments of the copying.

This section has so far sought to reveal the trends in copying and sūtra art driving

scripture transcription in the eleventh through thirteenth centuries in order to understand the

emergence of the inventive style of sūtra transcription embodied in the jeweled-stūpa mandalas.

However, the sūtra art of this period has been heavily criticized in past scholarship. Komatsu

Shigemi ponders the idea that from a present-day perspective, this period‘s religious atmosphere

appears bizarre and fanatical. 135 Much harsher evaluations have been leveled against what was

perceived as the decadence of the era and the decline of the religious practices and beliefs. For

example, Tsuji Zennosuke disparages the practice of quantity copying from the eleventh century

onward, such as that of the Buddhist canon, as a sign of the decline of the religion. He views this

type of quantity transcription as a sad indication of formulization, which inevitably leads to

135
Komatsu, ―Ichiji sanrei no shakyō,‖ 4.

49
decline and eventually the production of religious icons and art as a mere hobby. 136 Beyond the

focus on the production of quantity, he cites the trend in copying to add ‗twists‘ or elaborations

in the search for novel manifestations as further evidence of this decline. Claiming that these

novel ‗twists‘ are simply products of an overconcentration on design, he lists as proof the

majority of the examples discussed in this section.137 Komatsu Shigemi analyzes the practices

employed in sūtra transcription and concludes that faith itself at the close of the Heian period is

drastically formulized and lacking in any real sincerity. He points to the practice of ‗one

character, three bows‘ as an example of the diminishing of sincere faith and the corruption of

religious practices, presumably because the copyist has attempted to obfuscate his degraded faith

by intensive and seemingly pious copying strategies. 138 A further example of this argument in

older scholarship comes from Tanaka Kaidō. 139 Tanaka sees the importation of Song dynasty

printed scriptures and the lack of the Zen school‘s focus on textual sūtras as key factors in the

decline of copying practices. He views the mutable fashions of sūtra taste, as manifested by an

increased preference for the new printed scriptures during the Kamakura period, as reasons for

the decrease in sūtra transcriptions. He also points to the innovations in copying techniques such

as those seen in the project of Unkei as excessively baroque and lacking in earnestness of faith, a

trend he sees as continuing into the Muromachi period. Criticizing elaborate measures like the

search for pure water and ink not made of animal products and the replacement of animal hair for

brushes with bark from a willow as the creation of obstacles for the sake of enhanced merit,

Tanaka claims such methods reveal the absence of the true spirit of sūtra transcription and the

136
Tsuji, Nihon no bukkyō shi, 644.
137
Ibid., 644-70.
138
Komatsu, ―Ichiji sanrei no shakyō,‖ 4.
139
Tanaka, Nihon shakyō sokan, 30-4.

50
presence of narcissism. 140 Such judgments are rarely found in scholarship now. To criticize an

entire era‘s religious practices as devoid of real faith, as if this is easily ascertainable or even

plausibly posited without slipping into anachronism, and as an omen of the decline of religion

smacks of the same fallacious argument which contends that at certain times and places people

were not following the real Buddhism because practices were not always in accord with doctrine.

Conclusion

This chapter excavated some practical aspects of the jeweled-stūpa mandalas concerning their

origins, both in terms of stylistic precedence as well as the culture of copying in the eleventh

through thirteenth centuries out of which the mandalas emerged. By locating the foundations of

the paintings in early Chinese transcriptions and by situating the mandalas amongst other

inventive and novel sūtra art and copying practices at the time of the mandalas‘ first production,

the paintings become intelligible less as having materialized mysteriously and without

precedence for a brief time and more as a particular aspect of a system of sūtra transcription that

trended toward the innovative and extreme. This examination is not to diminish the mandalas‘

inventiveness but to reveal the context of their creation—they represent an apotheosis of general

efforts to creatively and laboriously transcribe sūtras, especially given their high levels of artistic

achievement. However, within the culture of copying at the close of the Heian period, the

jeweled-stūpa mandalas were indeed highly original in one particular aspect: their utter

imbrication of text with image was unprecedented in previous sūtra transcription projects, a

subject further explored in chapter six.

140
Ibid., 31.

51
 Chapter Three

The Jeweled-Stūpa Mandalas as Historical Objects:

Analyses of the Formal Qualities and Context

Introduction

In preparation for exploring the theoretical interpretations of the jeweled-stūpa mandalas as a

whole, this chapter closely examines the formal qualities and probable commission contexts of

the Chūsonji, Danzan Shrine, and Ryūhonji sets. Because the most is known about the Chūsonji

set, the section treating these mandalas begins with an examination of the most likely

circumstances for the mandalas‘ commission and symbolic function. In order to postulate a

production date for the Danzan Shrine and Ryūhonji sets, I analyze the technique, style, and the

possible commission context (in the case of the Danzan Shrine version) or contextual clues from

the records of restoration (for the Ryūhonji version). All three sections then explicate the visual

properties of the mandalas. Given the sumptuous and complicated construction of the jeweled-

stūpa mandalas, this visual analysis begins with an explanation of each set‘s process of

production. The mandalas are then analyzed in separate parts exploring the visual qualities of the

textual stūpa and the encircling narrative vignettes.

Set History: Chūsonji Mandalas1

Unlike the two other jeweled-stūpa mandala sets, the history and circumstances of commission

of the Chūsonji set are not so elusive, and their examination allows us to understand the more

1
For full images of each of the Chūsonji mandalas and details of each scroll‘s narrative vignettes, see Miya, Kinji
hōtō mandara, plates 1-73.

52
personal nature of the paintings‘ commission. 2 Located in the small but culturally sophisticated

northern outpost of Hiraizumi, the Ōshū Fujiwara fashioned political and cultural legitimacy

through the appropriation and localization of courtly symbols of authority and the insignia of

Buddhist mandate. Examples include sūtra copying, Jōchō 定朝-style sculptures,3 and most

important to this study, the northern Fujiwara‘s devotion to the Golden Light Sūtra and to the

patronage of the jeweled-stūpa mandalas. It is in this context of acculturation between Ainu

(indigenous culture of northern Honshū and Hokkaidō, and arguably the first inhabitants of

Japan4) and Kyoto that the jeweled-stūpa mandalas were commissioned and created. By situating

the mandalas within the tumultuous context of twelfth-century Hiraizumi, we can open a window

on the distinctive cultural amalgamation that was Hiraizumi under the rule of the Ōshū Fujiwara.

Close examination of the paintings reveal the concern of the northern rulers for legitimized

political authority, the mingling of Kyoto aristocratic and Emishi culture, and even the patron‘s

innermost salvific desires and anxieties.

Until the mid-twentieth century, the mandalas were stored in Chūsonji‘s benzaitendō 弁

財天堂 in ten zushi 厨子, or miniature shrines, designed in 1705 to house the paintings. 5 The

black lacquered boxes measure 165 x 67.2 x 15 cm. The simple exterior has two doors and a

gilded interior space enshrining the unfurled scrolls, which are stabilized by golden lotus-shaped

supports (蓮台座 rendai za). In the middle of the interior of the doors are two informative

2
For full images of each of the Chūsonji mandalas and details of each scroll‘s narrative vignettes, see Miya plates 1-
73.
3
By Jōchō-style sculptures, I refer to the style popularized by the sculptor, Jōchō (d. 1057), and his workshop in
which they used a multi-block carving technique known as yosegi zukuri 寄木造. They also popularized a new
canon of proportions, creating the appearance of youth, balance, and roundedness in the sculptures. The Amida of
1053 by Jōchō at the Byōdōin is a perfect example of this style of sculpture.
4
The exact relationship between the Emishi 蝦夷 and Ainu アイヌ is a highly debated topic. Some argue that
Emishi were Ainu, but no general consensus has been reached. For more on the controversy, see Yiengpruksawan,
Hiraizumi, 18-19.
5
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 265. The following information concerning the zushi
comes from this source.

53
inscriptions. The right side lists the sūtra title and volume number. The left side records an Edo

period categorical title for the paintings, jūkai hōtō e mandara 十界宝塔絵曼荼羅 or Ten World

Jeweled-Stūpa Mandala. Hamada Takashi characterizes the ‗ten worlds (十界 jūkai)‘ of the title

as a reference to the ten levels of the mandalas‘ stūpa—including the first story‘s false or pent

roof (裳階 mokoshi).6 Kameda Tsutomu advances a similar argument, explaining that the nine

floors plus the pent roof, collectively called jūkai 十階 or ten stories, came to be known as jūkai

十界, a phrase he notes is completely unrelated to the Golden Light Sūtra;7 presumably the

homonymic quality of the words is responsible for the transference. However, neither author

provides support for this supposition, and given the lack of textual records for the jeweled-stūpa

mandalas, perhaps it is equally as possible to suggest that the ‗ten worlds‘ refers to the ten scrolls

of the set, rather than to the ten stories of the stūpas, which is itself an inaccurate count.

Takahashi Tomio 高橋富雄 also finds this particular explanation weak and suggests instead that

jūkai 十界 refers to the number of scrolls, culminating in a statement about the transformation of

all things into the lands of the Buddha: one scroll, one stūpa, one world, and thus ten scrolls, ten

stūpas, and the worlds of the ten directions (十方世界 Jpn. jippō sekai, Ch. shifang shijie; Skt.

daśa dig loka dhātu), symbolizing the infinite expanse and all-encompassing nature of the

Buddha-realm. 8 Precisely because no data remains about the paintings, they have been given

multiple titles. Kameda also notes that the mandalas have been referred to as Kiyohira hōnō 清衡

奉納 or the ‗dedication by Kiyohira.‘9 As discussed below, Fujiwara Kiyohira 藤原清衡 (1056-

6
Ibid.
7
Kameda, ―Jūbun saishōōkyō jūkai hōtō mandara,‖ 68.
8
Takahashi Tomio 高橋富雄, ―Chūsonji to hokekyō: Chūsonji konryū no kokoro 中尊寺と法華経: 中尊寺建立の
心,‖ Tōhoku daigaku kyōyōbu kiyō 東北大学教養部紀要 33 (1981): 39.
9
Kameda, ―Jūbun saishōōkyō jūkai hōtō mandara,‖68.

54
1128) was the patriarch of the Ōshū Fujiwara, and research has shown that he was likely

unassociated with the production of the mandalas. Miya records that in1968 the mandalas were

officially registered as Konshi choshoku konkōmyōsaishōōkyō kinji hōtō mandara zu 紺紙著色

金光明最勝王経金字宝塔曼荼羅図 (a title that translates rather awkwardly into English as

Jeweled-Stūpa Mandala of the Golden Light Sūtra in Gold Letters on Blue Paper), thus

establishing the standardized title of the paintings. 10

Commission Context

As I have already mentioned in the discussion on the extreme lengths undertaken in sūtra

transcription projects, Hiraizumi during Ōshū Fujiwara rule rivaled the Kyoto court in artistic

commissions in terms of precious materials and the sheer scope of single projects. Documents

like Petition of the Bunji Era (文治の注文 Bunji no chūmon) composed in 1189 for Minamoto

Yoritomo 源頼朝 (1147-99) by the Chūsonji monks, Genchō and Shinren, offer a glimpse of

twelfth-century Hiraizumi and the extensive building campaigns of the Ōshū Fujiwara. 11 Because

of the breadth of patronage carried out during the three generations‘ governance, I must limit my

focus to sūtra transcription commissions. As Hamada Takashi points out, the Ōshū Fujiwara

during this time enjoyed great financial success allowing for expensive and laborious artistic

productions and, to aid in this endeavor, established a center for sūtra copying (写経機関 shakyō

kikan) known as Chūsonjikyō 中尊寺経.12 The celebrated sūtras known as the Kiyohirakyō 清衡

10
Miya, Kinji hōtō mandara, 34 n1.
11
Yiengpruksawan, Hiraizumi, 69. Kuroita Katsumi 黑板勝美, ed., ―Bunji no chūmon 文治の注文,‖ in Azuma
kagami 吾妻鏡, in Shintei zōho, kokushi taikei 新訂增補, 國史大系, vol. 32 (Tokyo: Yoshikawa Kōbunkan, 1929-
64), 352-55.
12
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 264.

55
経13 captured the entire Buddhist canon on blue paper with alternating lines of gold and silver

text (紺紙金銀字交書一切経 konshi kinginji kōsho issaikyō).14 This vast project was

commissioned by the patriarch of the Ōshū Fujiwara clan, Kiyohira, and dedicated in 1126. 15

Most of the scrolls have made their way under not so illustrious circumstances to Mt. Kōya‘s 高

野山 Kongōbuji 金剛峯寺 by the command of the powerful warlord, Toyotomi Hideyoshi 豊臣

秀吉 (1536-98). Kiyohira‘s son, Fujiwara Motohira 藤原基衡 (1105-57), and grandson,

Fujiwara Hidehira (1122-1187), continued the practice of elaborate sūtra transcription. Motohira

commissioned a set of ornate Lotus Sūtra scrolls, and Hidehira followed the tradition of his

grandfather and ordered a blue and gold Buddhist canon. 16 Hidehira‘s scrolls were enshrined at a

sūtra repository at Motohira‘s temple, Mōtsūji 毛越寺, and unlike the dispersal of Kiyohira‘s

copy of the Buddhist canon, most of the extant scrolls have remained at Chūsonji. 17 Given the

rarity of such sumptuous transcription projects as the blue and gold (and silver of Kiyohira‘s)

Buddhist canon, not to mention the many other sūtra commissions, copying the scriptures was an

important ritual conveying the Ōshū Fujiwara‘s political and salvific ambitions. Because of the

lack of documentary evidence locating the exact circumstances of the jeweled-stūpa mandala‘s

13
Although the exact temple consecration it refers to is debated, the controversial text known as the ―Chūsonji
rakkei kuyō ganmon 中尊寺落慶供養願文‖ mentions the commission of a blue paper Buddhist canon with
alternating lines of gold and silver script, which is a reference to the vast scriptural project of Kiyohira. Hiraizumi
Chōshi Hensan Iinkai 平泉町史編纂委員会, ed., ―Chūsonji rakkei kuyō ganmon,‖ in Hiraizumi chōshi 平泉町史,
vol. 1 (Tokyo: Zoku Gunsho Ruiju Kanseikai, 1985), 59-61. This three volume publication contains a wealth of
resources concerning the history of Hiraizumi.
14
For a discussion of the technique of this very unusual style of sūtra transcription, see Sasaki Hōsei 佐々木邦世,
―Kingin kōsho no tejun to kōfu 金銀交書の手順と工夫,‖ in Kenrantaru kyōten 絢爛たる経典, ed. Sato Shinji
(Tokyo: Heibonsha, 1983), 132-34. For more on the rarity of this style, see Ishida Mosaku, ed., Chūsonji 中尊寺
(Tokyo: Asahi Shinbunsha, 1959), 16.
15
Others suggest alternative dates. For instance Tanaka Kaidō suggests 1124. For his discussion of Kiyohirakyō, see
Tanaka, Nihon shakyō sōkan, 390-94.
16
Most scholars agree that this project is the product of Hidehira, but because of the vagueness of records and
inscriptions, others have suggested that Motohira originally commissioned the set and Hidehira completed it
sometime between 1150 and 1170. Part of the confusion arises from a postscript on the eighth scroll of the Lotus
Sūtra which testifies Hidehira‘s wish for the peaceful rest of his father. See Yiengpruksawan, Hiraizumi, 111.
17
Ibid.

56
commission, I follow the methods pursued by previous scholars in order to hypothesize the date

and patron of the paintings: by examining the contextual circumstances of Hiraizumi during the

rule of the three generations of Ōshū Fujiwara as well as by analyzing the styles of their artistic

commissions, a rough history of the mandalas may be sketched.

Mimi Yiengpruksawan, citing evidence gathered by Mori Kahei, 18 notes that before the

sixteenth century the only mention of the jeweled-stūpa mandalas was that they were

commissioned by Hidehira and stored in Chūsonji. 19 However, as the debate in the scholarship

shows, this does not settle the matter of the commission. Miya believes that such an extravagant

project warrants a pivotal event to commemorate.20 From his point of view, three key events in

twelfth-century Hiraizumi stand out: in 1126 under the direction of Kiyohira, Chūsonji held a

massive dedication ceremony; in 1170 Hidehira was promoted to the constabulary position of

‗pacification‘ general (鎮守府将軍 chinjufu shōgun); and in 1181 Hidehira was again promoted

to a position of great and independent power as the governor of Mutsu province (陸奥守 mutsu

no kami).21 Miya further notes that during this timeline, Motohira built Mōtsūji. 22 Because the

pledge associated with Kiyohira‘s dedication ceremony records many dedicatory objects yet

remains silent on the topic of the jeweled-stūpa mandalas, which surely would have warranted a

prominent place in the description of the services and commissions, Miya discounts Kiyohira as

a possible patron.23 He also rejects the ascension of Hidehira to the rank of Mutsu governor as

the likely event because the 1181 promotion occurred after the commission of Hidehira‘s

18
Iwate Nippōsha 岩手日報社, ed., Yomigaeru hihō Chūsonji konjikidō よみがえる秘宝: 中尊寺金色堂
(Morioka: Iwate Nippōsha, 1974), 41. Ishida, examining the temple records, remarks that they claim Hidehira
copied the mandalas. See Ishida, ―Kokuhō saishōōkyō kyōtō mandara,‖ 4.
19
Yiengpruksawan, Hiraizumi, 174.
20
Miya, Kinji hōtō mandara, 33. And as Yiengpruksawan notes, these powerful and high profile appointments did
not go without critical commentary by Kyoto contemporaries. See Yiengpruksawan, Hiraizumi, 97.
21
Miya, Kinji hōtō mandara, 33.
22
Ibid.
23
Ibid.

57
Buddhist canon in 1176. 24 The remaining possibilities for Miya then are either the ascension of

Motohira at the death of Kiyohira to the position of family patriarch and lower-level military

bureaucrat for Mutsu and Dewa 出羽 (押領使 ōryōshi) in 1128 (although this elevation was

consummated only after the murder of his older brother and rightful heir, Koretsune 惟常), or the

promotion of Hidehira to chinju shōgun in 1170, while still allowing for the possibility of

Hidehira‘s 1181 promotion. 25 And even though Miya finds the first two potentialities more

probable, the possibility persists that the mandalas are what he describes as ‗national products or

projects‘ (国家的事業 kokkateki jigyō). The elevated position of the rulers comes with access to

taxes for use in the construction of the paintings, and so the mandalas could make a statement not

only about the aristocracy of the Ōshū Fujiwara expressed through the manner and style of the

commission but also as a proclamation of their firm rule of the northern province. 26 In the end,

Miya seems to side with the 1170 date as the probable occasion. 27 Yiengpruksawan argues that

Hidehira‘s appointment to chinju shōgun in 1170 is the most likely occasion for the production

of the mandalas given the Golden Light Sūtra‘s strong message of righteous authoritarian rule. 28

Additionally, the ceremony for Hidehira‘s surprising elevation took place at the imperial palace

during the annual saishōkō 最勝講, an imperially sanctioned ceremony reaffirming the Golden

Light Sūtra as guardian of the nation and legitimizer of imperial authority, a symmetry that

Yiengpruksawan highlights as additional confirmation of Hidehira as the patron of the

paintings. 29

Other scholars advocate Motohira as the patron of the jeweled-stūpa mandalas. Hamada

24
Ibid.
25
Ibid.
26
Ibid.
27
Ibid., 122.
28
Yiengpruksawan, Hiraizumi, 174.
29
Ibid.

58
Takashi acknowledges that amongst the three generations, Motohira‘s era lacks the most

clarity. 30 Yet, when considering the Hiraizumi projects attributed to Motohira, such as the

elegant golden hall (金堂 kondō) of Mōtsūji; Kashōji 嘉祥寺, another temple built by Motohira,

and the paintings of the Lotus Sūtra decorating its walls; and Kanjizaiō-in 観自在王院, a temple

founded by Motohira‘s wife, and the landscape drawings ornamenting its walls, Hamada believes

that the artistic period under Motohira offers the greatest possibility for the commission of the

mandalas. He also suggests that given the mandalas‘ focus on righteously ordained power via the

choice of the Golden Light Sūtra (a topic addressed below), the more appropriate time for such a

subject would be earlier in the three generations‘ rule because the solidification of Ōshū power

occurred before Hidehira‘s era. 31 Hayashi On acknowledges the possibility of Hidehira‘s

patronage, recognizing that the 1170 promotion could be a triggering occasion, 32 but also makes

a case for the contextual plausibility of Motohira. Hayashi looks to the historical connectivity of

the Golden Light Sūtra and the Lotus Sūtra as national protective sūtras (護国教典 gokoku

kyōten) beginning with the 741 edict that all nationally sponsored temples (kokubunji) enshrine

ten copies of these two sūtras and chant both.33 By characterizing the sūtras in terms of strong

gender affiliations in which he views the Golden Light Sūtra as associated with the protection of

kings and the Lotus Sūtra connected with the plight of women, Hayashi portrays the two

scriptures as a husband and wife team working jointly to protect the nation. 34 Based on this

relationship, he points to Motohira‘s sūtra transcription project honoring Kiyohira (基衡願経

Motohira gankyō) by commissioning a section of the Lotus Sūtra to be copied once a day for one

30
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 264.
31
Ibid ., 265.
32
Hayashi, ―Daichōjuinzō konkōmyōsaishōōkyō kinji hōtō mandara zu oboegaki,‖ 82. For an additional discussion
of the strong relationship of the Lotus Sūtra with the Ōshū Fujiwara, see Takahashi, ―Chūsonji to hokekyō,‖ 19-41.
33
Hayashi, ―Daichōjuinzō konkōmyōsaishōōkyō kinji hōtō mandara zu oboegaki,‖ 93.
34
Ibid.

59
thousand days (千部一日経 senbu ichinichikyō).35 This extensive project began on the sixth day

of the eighth month of 1128 as a memorial service for Kiyohira‘s thirty-seventh death

anniversary. 36 By the seventh day of the sixth month in 1148 up to 572 pieces were made. 37

Hayashi ponders whether this commemorative occasion involving such a large-scale copying of

the Lotus Sūtra might also have included a corresponding and equally impressive sponsorship of

the Golden Light Sūtra.38

Examining the historical circumstances can lead to tentative claims of patronage and

dating. By comparing the stylistic qualities of the sūtra copies produced under all three Ōshū

Fujiwara, the subject is further illuminated. There is broad consensus that the style of the

jeweled-stūpa mandalas bares little similarity to the sūtras of Kiyohira. 39 Rather, a strong

resemblance is seen in the style of the arabesque design of peony, lotus, and other intertwining

flowers (宝相華唐草 hōsōge karakusa) from the Hidehira Buddhist canon and the arabesque

border framing the mandalas, although Hayashi describes the flowers of Hidehira‘s scrolls as

more formulaic and sees the pattern used in the front cover of Motohira‘s Lotus Sūtra as more

strongly related.40 Yiengpruksawan points out that the frequent motif of golden wheels peppering

the landscape of the mandalas‘ narrative vignettes correspond to the illustrations of the 1170-72

Taira Lotus Sūtra blue and gold scrolls dedicated to Itsukushima Shrine; indeed, this motif is

uncommon in earlier examples.41 While it is easy to see the direct similarities between the

mandalas‘ narrative vignettes and those occurring in twelfth-century blue and gold frontispieces
35
Ibid.
36
Ibid.
37
Ibid.
38
Ibid.
39
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 264; Hayashi, ―Daichōjuinzō konkōmyōsaishōōkyō
kinji hōtō mandara zu oboegaki,‖ 82; Kameda, ―Jūbun saishōōkyō jūkai hōtō mandara,‖ 68; and Miya, Kinji hōtō
mandara, 33.
40
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 264; Hayashi, ―Daichōjuinzō konkōmyōsaishōōkyō
kinji hōtō mandara zu oboegaki,‖ 82; and Miya, Kinji hōtō mandara, 31-33.
41
Yiengpruksawan, Hiraizumi, 174.

60
including those of Hiraizumi, given the close range of dates for the Ōshū sūtra transcriptions and

the general inability to consistently date blue and gold frontispieces because of the general

patternization of the imagery and styles, 42 the most reliable information about the commission

history of the mandalas comes from contextual indicators.

Overall, the most compelling case to be made is for Hidehira as patron, based on

contextual and stylistic factors. Stylistically, the jeweled-stūpa mandalas suggest a date later than

both Kiyohira and Motohira, and the formal qualities of the mandalas and Hidehira‘s scrolls

resemble one another. And as is discussed below, the choice of the Golden Light Sūtra—an

anomalous one given the rarity with which it was selected for this format in both China and

Japan according to the extant literature—further clarifies the circumstances of the jeweled-stūpa

mandalas‘ commission and points to Hidehira as the most likely patron.

Localized Mandalas and Goals

Because conclusive information about the commission or dating is impossible without firm

textual evidence, what we can safely interpret--given the patronage patterns of not only the Ōshū

Fujiwara but the broader practices of producing elaborate projects to memorialize and venerate

important public and personal events and dates--is that the jeweled-stūpa mandalas likely

functioned as a fantastic and profoundly personal copying project proclaiming the northern

Fujiwara rule while also revealing their interior anxieties. In this way, the initial choice of the

Golden Light Sūtra for the jeweled-stūpa mandalas is a revealing one. The Golden Light Sūtra

enjoyed significant imperial patronage beginning in the Nara period as a scripture capable of

protecting the state. For example, when Emperor Shōmu established the nationalized provincial

temple system, he ordered copies of the Golden Light Sūtra and the Lotus Sūtra to be enshrined
42
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 264.

61
within seven-storied stūpas at each of the outlying temples, thus blanketing his country with the

apotropaic scriptures.43 Critical to imperial ideological and political goals, the sūtra provided

support for political and theological power of the emperor with its discussion of the wheel-

turning king (轉輪聖王 Jpn. tenrin jōō, Ch. zhuanlun shengwang; Skt. cakravartin). Through

great virtue and sincere penitence, the wheel-turning king is divinely sanctioned, ruling his

empire in the name of the Buddha as the ultimate authority on earth and reaping great benefits

from his devotion to the sūtra. Such connotations of power were not lost on emperors, who

claimed the sūtra as a mandate. With divine support, the emperor, who conformed to the laws

established by the sūtra, ruled under the protection of the Four Guardian Kings:

Wherever, dear Lord, in future time this excellent Suvarṇabhāsa, king of sūtras, will go
forth in villages, cities, settlements, districts, lands, royal palaces, and whichever king of
men‘s region it may reach, whichever king of men, dear Lord, there may be who will
exercise sovereignty in accordance with the treatise on kingship (called) ‗Instruction
concerning Divine Kings‘, who will hear, reverence, honour this excellent Suvarṇabhāsa,
king of sūtras, and will respect, venerate, reverence, honour those monks, nuns, laymen
and laywomen who hold the chief sūtras and will continually listen to this excellent
Suvarṇabhāsa, king of sūtras, by this flowing water of the hearing of the Law and by the
nectar juice of the Law, he will magnify with great might these divine bodies of us four
great kings with our armies and retinues and those of the numerous hundreds of
thousands of Yakṣas. And he will produce in us great prowess, energy and power. He will
magnify our brilliance, glory and splendour. Therefore we, dear Lord, the four great
kings, with our armies and retinues and with numerous hundreds of thousands of Yakṣas,
with invisible bodies, now and in future time, wherever we come upon villages, cities,
settlements, districts, lands and royal palaces, there this excellent Suvarṇabhāsa, king of
sūtras, will go forth, and we will give protection, will give salvation, assistance, defence,
escape from punishment, escape from the sword, peace, welfare to their royal palaces,
their lands, and their regions. And we will deliver those regions from all fears,
oppressions, (and) troubles. And we will turn back foreign armies. 44

The sūtra was read annually at the saishōkō and was a centerpiece of the annual misaie 微細会, a

43
Brian D. Ruppert, Jewel in the Ashes: Buddha Relics and Power in Early Medieval Japan (Cambridge, MA:
Harvard University Asia Center, 2000), 60.
44
R.E. Emmerick, trans., The Sūtra of Golden Light, 3rd ed., (Oxford: Pali Text Society, 1996), 27. T. no. 665, 16:
427c7-20. Emmerick‘s translation from the Sanskrit is not exact to Yijing‘s text, but when his English translation is
used in this project, the main ideas and general structure and wording of the passages are similar.

62
ritual for the protection of the emperor and his rule.45 Indeed, annual saishōkō were held at

Chūsonji and Mōtsūji and several Golden Light Sūtra copies with gold letters were stored in the

sūtra repository of Chūsonji, highlighting the Ōshū Fujiwara‘s devotion to this sūtra. 46

Key themes stressed in the narrative vignettes reveal particular inclinations and motives

on the part of the patron. Probably the most prominent and consistently featured motif in the

pictorializations is the Four Guardian Kings, including the lone figure of Bishamonten (毘沙門

天 Ch. Pishamentian; Skt. Vaiśravaṇa). Indeed, the guardian kings appear in six of the ten

scrolls 47 in a distinctive iconographical style in which they are the only inhabitants of the

jeweled-stūpa mandala realm rendered in fine, gold, outline-style drawing. The visual

prominence of the guardian kings mirrors the critical and active role that the deities play in the

Golden Light Sūtra. Significant passages are dedicated to extolling the Four Guardian Kings‘ and

other tutelary deities‘ protection for those who hold and keep the sūtra; specifically, the twelfth

chapter of Yijing‘s translation of the sūtra, The Protection of the Nation by the Four Guardian

Kings,48 details the vast rewards offered to those–and in particular, kings and monks–who revere

the sūtra. The chapter begins with the promise of protection from encroaching enemies, freedom

from sundry afflictions, and salvation from the bitterness of famine and epidemics for those who

follow the Golden Light Sūtra.49 The Four Guardian Kings swear an oath to smite and subdue

oppressors and to destroy evil and disease by the great power and authority bestowed upon them

as defenders of the righteous followers of the scripture.50 The promises of such sought-after

blessings often focus on the eradication of enemies, devoting long passages of strong rhetoric

45
Ruppert, Jewel in the Ashes, 103-07.
46
Miya, Kinji hōtō mandara, 33.
47
The figures appear in scrolls two, three, five, six, seven, and eight. In scroll eight, Bishamonten is the only figure
represented from the group.
48
Jpn. Shitennō gokoku bon, Ch. Sitianwang huguo pin; 四天王護國品; T. no. 665, 16: 427b20-432c10.
49
T. no. 665, 16: 427c1-6.
50
T. no. 665, 16: 427c9-28.

63
detailing the utter annihilation of adversaries and their lands:

If there should be another hostile king neighbouring upon that king of men who hears,
reverences, honours this excellent Suvarṇabhāsa, king of sūtras, and if, dear Lord, this
neighbouring hostile king should produce such a thought: ‗I will enter that region with a
four-fold army to destroy it,‘ then indeed, dear Lord, at that time, at that moment, by the
power of the brilliance of that excellent Suvarṇabhāsa, king of sūtras, there will arise a
conflict between that neighbouring hostile king and other kings. And there will be
regional disturbances in his own regions. There will be fierce troubles with kings, and
diseases caused by planets will become manifest in his area. Hundreds of different
distractions will become manifest in his area. And if, dear Lord, there should arise for
that neighbouring hostile king in his own area hundreds of such various oppressions and
hundreds of various distractions, and (if), dear Lord, that neighbouring hostile king
should employ his fourfold army to go against a foreign power and it should leave his
own area, and (if) that hostile king together with his fourfold army should desire to enter,
should desire to destroy that region where this excellent Suvarṇabhāsa, king of sūtras,
may be, we, dear Lord, the four great kings, with our armies and retinues, with numerous
hundreds of thousands of Yakṣas, with invisible bodies, will go there. We will turn back
that foreign army from the very path it has taken. We will bring upon it hundreds of
different distractions, and we will make obstacles so that that foreign army will not be
able to enter this region, much less cause destruction to the region. 51

Perhaps this emphasis would have been of comfort to the patron of the mandalas because

of the tenuous relationship with the Kyoto court and with the Minamoto clan, a peace ultimately

broken with the devastating destruction of Hiraizumi in 1189 during the war between Minamoto

Yoritomo and the remaining Ōshū Fujiwara. Thus, the choice of the Golden Light Sūtra by the

most probable patron, Fujiwara Hidehira, is significant for exploring the purpose of the mandalas.

Because of the scripture‘s political and ideological symbolism, the commissioning of the

paintings by a northern war lord is laden with implications. Hidehira likely commissioned the

mandalas soon after his appointment as chinju shōgun in 1170, linking the mandalas with the

claim of legitimacy for a northern, holy rule. Interestingly, while Chinese precedents exist, the

Chūsonji paintings represent the first use in Japan of the Golden Light Sūtra to construct the

central stūpa of the mandalas. Possibly, in their quest for legitimacy, the northern Fujiwara

circumvented Kyoto altogether in favor of continental precedents. What is substantiated is that

51
Emmerick, The Sūtra of Golden Light, 27-28. T. no. 665, 16: 427c20-7.

64
the sūtra‘s efficacious ability to secure political authority via divine sanction offered an

enormous appeal to the northern rulers. Thus, the jeweled-stūpa mandalas espouse a manifesto of

righteously ordained power through their devotion to the Golden Light Sūtra, a sūtra traditionally

employed for the protection of the emperor and his rule. The decision to adapt a sūtra so closely

identified with the conventional sources of power by northern rulers whose authority was

continually contested marks an attempt at legitimacy and recognition through the sumptuous and

rare format of the Golden Light Sūtra jeweled-stūpa mandala, but it also reveals the tenuous

nature of their rule in its prayers for persistent protection.

The strong faith in the Four Guardian Kings, manifested through the visual dominance of

the deities in the mandalas‘ vignettes, reflects the Tōhoku 東北 area‘s belief in the guardians in

general and in Bishamonten in particular.52 The images of the Four Guardian Kings serve as

visual prayers for heavenly protection and investment of divine authority, a request also seen in

Kiyohira‘s pledge at the time of his tremendous donation of sūtras, sculptures, and stūpas, among

other objects, in 1126 to save his realm: ―奉建立供養鎮護国家大伽藍一区 Hōkonryu kuyō

chingo kokka daigaran ikku.‖53 As Hayashi On indicates, the Ōshū Fujiwara‘s awareness of their

uneasy position of autonomous authority over the north was well-engrained. 54 Through a

comparison of the conventional iconographic positions of the Four Guardian Kings with the

ordering of the guardians in the Chūsonji mandalas, Hayashi interprets this atypical alignment as

the visual claim of the Ōshū Fujiwara‘s autonomous authority over northern Honshū.55 When

arranged in a three-dimensional or stacked configuration, the characteristic allocation takes the

form of Jikokuten 持国天 (Ch. Chiguotian; Skt. Dhṛtarāṣṭra) in the lower right position marking

52
Miya, Kinji hōtō mandara, 23-26.
53
Hayashi, ―Daichōjuinzō konkōmyōsaishōōkyō kinji hōtō mandara zu oboegaki,‖ 92.
54
Ibid., 92.
55
Ibid., 87.

65
east, Zōchōten 増長天 (Ch. Zengzhangtian; Skt. Virūḍhaka) in the lower left position marking

west, Kōmokuten 広目天 (Ch. Guangmutian; Skt. Virūpākṣa) in the upper left position marking

west, and Tamonten 多門天 (Ch. Duowentian; Skt. Vaiśravaṇa) in the upper right position

marking north.56 But as Hayashi observes, the standard arrangement of the guardians in the

jeweled-stūpa mandalas flips the north-south axis, switching the positions of Tamonten and

Zōchōten in all but the third scroll where Kōmokuten assumes Zōchōten‘s NE position. 57

Hayashi infers from this switch, along with the resulting emphasis on the north (Tamonten) and

east (Jikokuten) achieved by maneuvering them into the frontal positions, that the perpetually

directionally-conscious patron of the mandalas asserted the dominance of the NE, the geographic

position of Hiraizumi in relation to Kyoto.58

Rather than the wholesale adoption of the Kyoto trappings of culture and legitimacy

resulting in the jettisoning of Emishi culture, the Ōshū Fujiwara transformed Hiraizumi while

maintaining traditions and symbols important to their Emishi heritage. The Chūsonji mandalas

manifest this attitude in the treatment of the landscape, scenes of excessive violence, and

prominence of women in the paintings, revealing the signature character of Hiraizumi, its war

lords, and its traditional customs. As Yiengpruksawan observes, in the frontispieces of the blue

and gold type of sūtra transcription of the Chūsonjikyō and in the narrative vignettes of the

mandala the rendered landscape reveals similarities to the actual terrain of the Hiraizumi area. 59

By localizing the vignettes of the frontispieces and mandalas, a more personal and intimate

association with the sūtra is established. And much like the sūtra frontispieces, scenes of

excessive violence—extraneous to the scriptural content—populate the mandalas. In the

56
Ibid..
57
Ibid., 83, 87.
58
Ibid., 87.
59
Yiengpruksawan, Hiraizumi, 117, 177.

66
frontispieces of Hidehira‘s Buddhist canon, unexplained and unconnected images of grave

violence are depicted. For instance, before the observant eyes of the Buddha and his attendants, a

monk stretches taut the string of his bow, directing his arrow toward a flock of ducks. In another

frontispiece, demons gleefully mutilate a person, shoving the body headfirst into a meat grinder

while the emaciated hungry ghosts (餓鬼 Jpn. gaki, Ch. egui; Skt. preta), with their painfully

distended bellies, attempt to slake their sufferings with food and drink. In the third scroll of the

jeweled-stūpa mandalas, a body lays decapitated and spilling blood while above it a man quickly

prepares to pierce it with a third arrow, all while Jizō watches. Adjoining that scene, a figure

swings the axe that will inevitably fell a Buddha image enshrined in a free standing structure.

Thus this vignette describes two of the most tragic sins: murder and crimes against

Buddhism. Yiengpruksawan postulates that the Ōshū Fujiwara‘s preoccupation with violence in

the form of hunting, war, and torture reveals not only Emishi cultural traditions and ways of life

but also anxieties over salvation rife in the generations of northern war lords. 60 She suggests that

giving visual utterance to such violence might be the patron‘s adherence to the commands of the

Golden Light Sūtra which call upon all rulers who wish to be successful to confess and repent

their karmic transgressions.61 Indeed, the third scroll in this set visualizes the fifth chapter of the

scripture, the chapter dedicated to confession and, through penitence, salvation. The sūtra

cautions against the dangers of ignorance and warns that in times of estrangement when one does

not know the Buddha, dharma, monks, or even good and bad then one is perilously close to

committing endless crimes because in such a state of dismal ignorance, one is incapable of

discerning right from wrong. Therefore, these evils perpetrated, even unknowingly, result in

injuring the body of the Buddha, usurping justice, destroying the harmony of the monastic

60
Yiengpruksawan, Hiraizumi, 118-20.
61
Ibid., 176.

67
community, and killing arhats (羅漢 Jpn. rakan, Ch. luohan) and even parents.62 A proper

confession of crimes and ignorance is required:

In the oppression of existence (or) through foolish thought, whatever severe evil I have
done, in the presence of the Buddha, I confess all this evil. And I confess that evil which
has been heaped up by me in the oppression of birth, by the various oppressions of
activity due to passion, in the oppression of existence, in the oppression of the world, in
the oppression of the fleeting mind, in the oppression of impurities caused by the foolish
and stupid, and in the oppression of the arrival of evil friends, in the oppression of fear, in
the oppression of passion, in the oppression of hatred and in the oppression of folly and
darkness, in the oppression of the opportunity, in the oppression of time, in the
oppression of gaining merits, standing in (my) oppression before the Buddha, I confess
all this evil. 63

It is therefore only appropriate that this confessional scene would appear on the third scroll

devoted to the admission and exoneration of sins.

Additionally, the figure of Jizō in the vignette of violence is wholly unconnected with the

sūtra, raising questions about his placement in this brutal scene. The insertion of Jizō into the

mandala represents another example of the personalization of the set conveying the desires and

fears of the Ōshū Fujiwara. Jizō‘s presence in a project with which he is scripturally unaffiliated

exposes the private relationship the patron experienced with the deity so celebrated for his

salvific abilities. Jizō‘s willingness to assume multiple manifestations in order to intervene on

behalf of sinners, to enter even the depths of hell, made him a very popular figure in the Heian

period and a particularly poignant one for the Ōshū Fujiwara. Hayashi observes that even though

Jizō is a bosatsu, he is not painted in gold as the other high-ranking deities but instead is

rendered in color like the people who populate the realm of the mandalas. 64 This visual nod to

Jizō‘s humanity affirms his intercessional proclivities. Jizō is given a very prominent role in the

konjikidō 金色堂 of Chūsonji; here he manifests as the Six Jizō (六地蔵 Roku Jizō) whose

62
T. no. 665, 16: 414a15-18.
63
Emmerick, The Sūtra of Golden Light, 13. For a long passage containing many expressions of repentance and
confession, and the absolution of those sins, see T. no. 665, 16: 412a27- b22.
64
Hayashi, ―Daichōjuinzō konkōmyōsaishōōkyō kinji hōtō mandara zu oboegaki,‖ 89.

68
dedicated function is to search out and save sinners trapped in the Six Paths (六道 rokudō) of

suffering. This efficacious materialization is a part of a larger, singular program visualizing an

elaborate and costly prayer for redemption and salvation in the form of a family mausoleum (墓

堂 kidō). This small hall is coated with black lacquer and covered with gold leaf on both the

interior and exterior. Precious gems, stones, shells, pearls, and other exquisite materials decorate

the radiant walls that serve to entomb the bodies of the three generations of northern Fujiwara. 65

Each of the altars displays an involved program of Six Jizō stationed in two rows of three along

the sides of the altar, an Amida triad (阿弥陀三尊 Amida sanzon) in which Kannon (観音 Ch.

Guanyin; Skt. Avalokiteśvara) and Seishi (勢至 Ch. Shizhi; Skt. Mahāsthāmaprāpta) flank the

central Amida, and pairs of guardians—Zōchōten and Jikokuten—posted along the front of the

altar. The anomaly of a hall that mixes so intimately the sacred with the profane, the pure and

eternal with the decaying and fleeting—despite the attempts for immortal preservation—is

captured by Yiengpruksawan: ―If scholars have often argued about Konjikidō or been puzzled by

its ambiguities, it is because the hall contains the story of the Hiraizumi Fujiwara in the privacy

of their death.‖66 The anxiety over sins committed, the disquiet of approaching death, and the

fear that paradise is beyond reach that permeate the visual program of the konjikidō and the

mummification of their bodies is also embodied in the visual program of the jeweled-stūpa

mandalas.67 Repentance for rewards so stressed by the Golden Light Sūtra is not only aimed at

political ambitions in the context of the mandalas, but soteriological ones as well.

The scene of bloodshed and Jizō is not the only vignette connected with this driving

65
The head of Yasuhira of the fourth generation of Ōshū Fujiwara also found its final resting place in the altar of his
father, Hidehira, making the entombment actually three bodies and one head. For more information about this
spectacular and complex hall, see Yiengpruksawan, Hiraizumi, 121-160.
66
Ibid., 137.
67
For a thorough discussion of the konjikidō mummies, see Ishida, Chūsonji.

69
desire. Throughout the mandalas, the motifs of descending musical bosatsu (奏楽菩薩 sōgaku

bosatsu) and Amida triads, again with the conventional attendants Kannon and Seishi, recall the

promise and visual culture associated with Pure Land Buddhism (浄土宗 Jpn. Jōdoshū, Ch.

Jingtuzong). The musical bosatsu typically play an assortment of instruments while flower petals

fall around them, visualizing the sounds and scents of paradise come to earth. In the right left

register of the first scroll, nine bosatsu fly closely together at a steep angle implying a rapid

descent. The scene captures a moment of joyous, melodic celebration. The second grouping of

these figures occurs in the left middle register of the third scroll where this time ten bosatsu

gather. The clouds upon which they sit trail behind them in slender wisps, amplifying the speed

of the descent, and flower petals rain down around them. As Hayashi notices, these collections of

descending bosatsu playing musical instruments evoke familiar scenes of the Welcoming

Descent (来迎 Jpn. raigō, Ch. laiying) associated with faith in Amida and his promise of

salvation. 68 The Amida triad found on the third scroll furthers testifies to this reoccurring desire

and anxiety. Neither the musical bosatsu nor the Amida triad relate to the sūtra content; instead,

these images should be seen in a similar vein as the scene of violence, the localized landscape,

and the prominence of women, discussed below. Once again, the prayers for the absolution of

sins and admittance into heaven as well as the Ōshū Fujiwara‘s syncretic faith incorporating Pure

Land Buddhism are revealed. The representation of blended and localized faith makes it

impossible to categorize the mandalas within one school of Buddhism, and Miya explains that

the reason for such diverse references and for the many vignettes unconnected to the sūtra

content is that the mandalas cater to a limited and specific audience to whom the content and

68
Hayashi, ―Daichōjuinzō konkōmyōsaishōōkyō kinji hōtō mandara zu oboegaki,‖ 88-89.

70
messages would be intelligible. 69

The inclusion of so many women unrelated to the scripture‘s subject matter in the

vignettes surrounding the stūpa also speaks to Emishi culture. The prominence of women in the

funeral traditions of the Emishi (and thus their ties to notions of salvation—not as conduits but

rather as facilitators who prepare the body for the funeral in procedures such as mummification),

their importance to the political aspirations of the Ōshū Fujiwara, and their active role in the

Buddhist pursuits at Hiraizumi are compelling reasons behind the inclusion of its many female

worshippers. Elizabeth ten Grotenhuis even suggests that the women depicted could be portraits

of Hiraizumi‘s women,70 and Ariga Yoshitaka argues that there is a strong possibility that the

women of the Ōshū Fujiwara were intimately connected to the making of this mandala, as they

were with numerous other artistic projects. 71 Therefore, the jeweled-stūpa mandalas of Chūsonji

manifest an intimate and personalized commission involving prayers for political authority

through the choice of the Golden Light Sūtra; images that reflect the Tōhoku area‘s regional

emphasis in Bishamonten worship;72 directional assertiveness through the manipulated alignment

of the Four Guardian Kings; the sūtra‘s focus on repentance and forgiveness, along with the

accompanying narrative vignettes, expressing the longing for the absolution of sins; and the

prominent focus on women in the narrative vignettes, visualizing the respected and involved role

of women in Emishi culture and perhaps even in the production of the mandalas.

Formal Analysis

Process of Construction

69
Miya, Kinji hōtō mandara, 26.
70
Quoted in Yiengpruksawan, Hiraizumi, 177.
71
Ariga, ―Konkōmyō saishōō kyō zu saikō,‖ 97-98.
72
For more on the prevalence of Bishamonten worship and imagery in Tōhoku, see Yiengpruksawan, Hiraizumi, 42-
47.

71
The enormity of the jeweled-stūpa mandala project at Chūsonji required precise and careful

planning for successful execution. Once again the lack of records means large lacunae of

information surround the actual copying of the mandalas, leaving the matter of the workshop and

artists obscured. However, through formal analysis, certain insights into the production of the

paintings and their place at the close of the Heian period, clearly a particularly tumultuous time

for the Ōshū Fujiwara, can be ventured. The strong contrast of the brilliant blue background,

shining golden characters, and vividly colored vignettes of blue-green style landscape with red,

indigo, white, silver, cinnabar, and deep green details of the figures and architecture compose an

arresting image, and as Ariga Yoshitaka reveals, a most unusual one. Ariga postulates that the

remarkable combination of gold characters copied onto a blue background with the new addition

of colorful vignettes developed in Hiraizumi as a result of the visual conflation of Kyoto‘s

conventional blue and gold illuminated sūtra transcription with the purple paper sūtra

transcription style which did incorporate multiple colors, such as the example seen in the

Kagawa Museum‘s copy of the Lotus Sūtra.73 Ariga observes that traditionally, colors do not

accompany the blue and gold (and occasionally silver) transcriptions, whereas when purple paper

is used, colors are added. The jeweled-stūpa mandalas, already an innovative transcription

project, display additional inventiveness in this merged approach to the visuals of sūtra copying.

Kameda Tsutomu ventures to call this contrasting and combinatory style the nature of

―Hiraizumi art.‖74 In an interesting proposal, Machida Seishi 町田誠之 advances the idea that

the blue that so often serves as the ground of sūtra copies has a more practical purpose: the

prevention of deterioration from insects and bacteria, but why this might be is left unstated. 75

And as he notes, using black ink upon a deep blue background would make the characters nearly
73
Ariga, ―Konkōmyō saishōō kyō zu saikō,‖ 93-98.
74
Kameda, ―Jūbun saishōōkyō jūkai hōtō mandara,‖ 68.
75
Machida Seishi 町田誠之, ―Mukashi no shōsokukyō 昔の消息経,‖ Nihon rekishi 日本歴史 584 (1997): 32.

72
illegible, and so gold ink is the contrasting partner needed.76

Miya proposes that the mandalas were originally free hanging scrolls (掛幅装 kakefusō)

now framed, 77 but Sudō Hirotoshi 須藤弘敏 and Iwasa Mitsuharu 岩佐光晴 believe that the

paintings were attached to the panels of a folding screen (屏風 byōbu) before they were removed

and framed. 78 The remarkably similar format and size, including the shape of the textual stūpa,

suggests that a pattern was used as a model for all ten paintings. Seven pieces of paper join

together to create the colorful world of the jeweled-stūpa mandalas. The central stūpa is

composed of two sheets of paper that meet at the fourth story, not including the first floor‘s pent

roof. The upper sheet reaches all the way to the top border of beautiful brocade flowers, upon

which is placed a small, separate piece of paper listing the title of the sūtra and the volume

number. The lower sheet terminates just under the handrail of the stairs. The sheets are closely

cropped to the shape of the stūpa, leaving very little extra space and implying careful planning of

the exact form of the reliquary before the dyeing and cutting of the paper. The narrative vignettes

are divided amongst five different sheets of paper. The bottom sheet spans the width of the

mandala beginning again just under the handrail, with the two sheets for the narrative vignettes

joining atop it on the right and left. The slender sheets used for the graphic scenes bordering the

stūpa, two per side, join at that same place directly outward from the fourth story of the stūpa.

The seams of the sheets carrying the narrative vignettes often coincide with vertical details of

landscape and architecture, effectively hiding the intersection of papers, while the vertical rows

of characters above the architectural brackets obscure the seam in the stūpa. The color

consistency of the stūpa‘s two sheets, made stark by their contrast with the hues often seen with

76
Ibid., 32.
77
Miya, Kinji hōtō mandara, 11.
78
Sudō Hirotoshi 須藤弘敏 and Iwasa Mitsuharu 岩佐光晴, Chūsonji to Mōtsūji 中尊寺と毛越寺 (Tokyo:
Hoikusha 保育社, 1989), 148.

73
the narrative vignette papers, implies that they were dyed at the same time. This observation

offers a sense of the sequencing, or at least, of the division of labor in the creation of the

mandalas. As has already been said, the overall pattern was well-established before any paper

was treated or cut. A formal analysis of the paintings suggests that the stūpa form was executed

first in the sequence of production. This idea is supported by the entrance of the narrative

vignette edges into the space of the stūpa, many times encroaching quite close to the stūpa, and

thus necessitating that the stūpa be finalized before the narrative vignettes were completed. This

would also seem to imply that the stūpa was executed before the vignettes were begun, but given

that a pattern was most likely used, contemporaneous production is possible with the final

touches to the scenes being added later. It is possible that the surrounding narratives were not

painted until all the papers of the mandala were joined, although this seems unlikely given the

extra care such a sequence would require. However, Hamada Takashi does seem to suggest this

approach. He explains that all ten textual stūpas were copied at the same time, which is a

convincing point, but then he offers the surrounding paper was attached and the chapter titles and

scriptural passages along with the narrative vignettes were then executed. 79 But as Hamada also

points out, certainly different artists were commissioned for the copying of the scripture and for

the painting of the surrounding scenes, as artists typically specialized in one or the other. 80

Stūpa

The central stūpa of each mandala contains the crucial architectural elements and crafts a picture

of an architecturally accurate reliquary, this time made of relic. The finial includes the seven

appropriate attributes: the crowning jewel (宝珠 hōju), the oval detail (竜車 ryūsha) between the

79
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 263.
80
Ibid.

74
jewel and the metal decoration affixed to the exposed central pillar (水煙 suien), the nine metal

rings attached to the central pillar (九輪 kurin), a simple flower base upon which the upper

structure rests (受花 ukebana), the inverted bowl which acts as a support to the flower base (伏

鉢 fukubachi), and finally, the square box maintaining the entire finial (露盤 roban).81 A simple

form of rafter known as a base rafter (垂木 taruki), the bracket arm that connects the beams to

the pillars (肘木 hijiki), and the bearing blocks for the bracket structure (斗 masu) craft each

floor.82 The latticing of the windows (連子窓 renjimado) on each floor is rendered in a faint

green and rimmed in a bright red square. The foundation (基壇 kidan) rests upon an inverted

lotus base (反花 kaeribana) with a banister (高欄 kōran) framing the foundation leading into a

handrail that descends along the stairs and flairs out at the bottom. The platform of the

foundation is colored a blue-green matching that of the surrounding hills and spits of land. The

once luminous silver of the paneling (腰板 koshita) facing out on the first floor has oxidized,

changing its color to a duller gray. Revealed between the opened doors of the first floor is a

seated golden Buddha offering the preaching mūdra (説法印 Jpn. seppōin, Ch. shuofa yin; Skt.

dharmacakra mudrā).

Amazingly, all of these architecturally accurate details are built of diminutive sūtra

characters. Given the demands of such a vast copying project, a pattern ensured consistency of

shape and size across all ten mandalas, and from close scrutiny, the grooved marks left by an

iron stylus are visible, revealing that the many lines of text composing the stūpa were

81
Miya, Kinji hōtō mandara, 12. For a diagram of a conventional stūpa finial, see ―Sōrin,‖ Japanese Architecture
and Art Net Users System, accessed January 14, 2011, http://www.aisf.or.jp/~jaanus/.
82
Miya, Kinji hōtō mandara, 12.

75
predetermined.83 The transcription of the Golden Light Sūtra‘s thirty-one chapters in ten volumes

begins from the uppermost, central character of the finial of each stūpa. From this peak point, the

sūtra continues in order down the nine-storied stūpa and finishes with the description of the last

post of the right-side handrail as it flares out from the bottom of the foundation‘s stairs. 84 Unlike

the Danzan Shrine and Ryūhonji sets that are precisely calculated to close with the end of each

volume, the Chūsonji mandalas are not adjusted to finish each time with the end of the volume.

Sometimes the concluding point is reached before the volume finishes.85 Other times, when the

volume closes before the stūpa‘s terminal point, the copyist simply iterates the process until the

stūpa is complete, reflecting a much more relaxed copying style in Chūsonji compared with the

other sets. As described in chapter one‘s visual introduction to the jeweled-stūpa mandalas, the

characters can appear in tidy, vertical lines while at other times, they turn on their sides to form

horizontal lines and even twist and dangle off the eaves of the roof. Legibility is sacrificed for

the integrity of the stūpa shape. By comparing the writing style of the characters building the

stūpa with that of the title and scripture excerpts accompanying the narrative vignettes, Kameda

has argued that the writing of the stūpa is of a more professional quality, while the texts

identifying the pictorial scenes show rougher and more simplistic brushwork. 86 From this

observation, he suggests that the titles and excerpts could have been brushed by the patron

himself. 87 Ishida Mosaku, however, disagrees with this assessment, finding that the characters of

the stūpa expose a consistency across the ten mandalas‘ stūpas indicating that the massive project

was performed by one person, but not the hand of a professional. 88 However, Kameda‘s

83
Miya, Kinji hōtō mandara, 119 and Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 262.
84
Miya, Kinji hōtō mandara, 119.
85
Ishida, ―Kokuhō saishōōkyō kyōtō mandara,‖ 5.
86
Kameda, ―Jūbun saishōōkyō jūkai hōtō mandara,‖ 68.
87
Ibid.
88
Ishida, ―Kokuhō saishōōkyō kyōtō mandara,‖ 4.

76
argument is the most convincing based on a visual analysis of the stūpas and cartouches. The

writing composing the stūpa and that of the surrounding cartouches do differ, implying different

brushes, and, it is difficult to think that the difficult task of copying the scripture would be given

to only one copyist. Rather, it seems likely that the studio of professional artists at work in

Hiraizumi cultivated the consistent style that we see in all ten stūpas.

Narrative Vignettes

Given the scale of the project, a close, visual analysis of the narrative vignettes of each scroll in

all three sets is infeasible. Therefore, my analysis is selected to tease out the overarching visual

themes and briefly highlight the key components of each scroll. Overall, this set depicts more

superfluous and uninhabited landscape serving as fillers and transitions between the vignettes,

and the total number of narrative scenes offered number fewer than either of the other two sets,

creating a more balanced and sparse composition with extensive negative space. A particularly

good example of the minimalistic compositions of the Chūsonji set is the sparse fascicle ten,

whose copious amounts of empty blue paper rival the space devoted to graphic depiction.

Whereas the Chūsonji mandala offer almost equal space to the graphic and negative areas,

Ryūhonji‘s mandala—albeit wider by approximately five centimeters—contains many more

vignettes, with traces of landscape and trails of clouds even extending into the areas between the

roof eaves, leaving little room for the blue paper to stand alone. As a near polar opposite of the

Chūsonji version, the Danzan Shrine mandala verges on confusion with its crowded composition.

The general configuration of the Chūsonji scenes begins in the upper right and moves down into

the space beside the foundation and then continues on the left side, moving again from top to

77
bottom, an allocation described as quintessentially Japanese. 89

The focus of the mandalas on the depiction of deities rather than narratives is a reflection

of the content of the sūtra, which does not lend itself as easily as the Lotus Sūtra to narrative

accounts. Indeed many of the scenes are assemblies of deities, representations of lectures, and

proclamations of the sūtra. In this way, many lack narrative content but instead manifest the

ideological and symbolic. Cartouches are not often supplied, but when they are, they frequently

only identify the chapter title referenced by the scene or detail the great benefits bequeathed to

those who venerate the sūtra and its followers. Actual passages from the scripture are quite rare.

Sometimes cartouches are present when the graphic illustration is absent. Miya finds indications

of general Heian sensibilities in the selection of the narrative vignettes, comparing the

ambiguities of the scenes with the vagueness and indirectness of Heian period literature, which

merely hint at meanings that are rarely explicitly discussed. 90 He elaborates that Heian-period

aristocrats expected of their educated contemporaries an erudition and competency in sūtra

content; therefore, ambiguous vignettes provided an opportunity to solve their meaning and

connect them to the appropriate scriptural passage. 91 However, this style in which representations

are often disconnected from scriptural content is explained by Hamada as either that the person

overseeing the production struggled to select which passages to pictorialize or that he lacked a

firm understanding of the Golden Light Sūtra.92 He continues, noting that this disconnect from

the scriptural content experienced by the vignettes could only have happened at a far-flung place

like Hiraizumi, so removed from the capital and thus freed from restrictions. 93 Hayashi

89
Miya, Kinji hōtō mandara, 120-21. This is in juxtaposition to the sinicized arrangement of Danzan Shrine‘s
narrative vignettes.
90
Ibid., 27.
91
Miya, ―Kenrantaru kyōten,‖ 97.
92
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 263.
93
Ibid., 264.

78
concludes that the painter of the vignettes could not have been a trained Buddhist painter.94

Rather than interpret the fluidity and ambiguity of the narrative vignettes and cartouches as either

mistakes or lack of understanding or training, it seems much more likely that these attributes are

further indications of the localization of mandalas intended for a specific audience with a specific

and tailored message.

Of course, there are scenes that are easily identifiable and clearly represent stories from

the sūtra. On the left side of the first scroll, the middle vignette describes the solemnity of

Myōdō Bodhisattva (妙憧菩薩 Ch. Miaochong pusa) before the assembled Buddhas as he asks

questions on the nature of the Buddha‘s lifespan from the chapter, ―The Longevity of the

Tathāgata.‖95 As he poses questions on the seemingly short life of Śākyamuni, four Buddhas

materialize in an ornamental building, stationed on elaborate lotus pedestals. The scene in the

mandala is rendered by a red, ornamental building enshrining five Buddhas, rather than the four.

Myōdō bosatsu sits on a golden lotus pedestal in supplication outside the structure. The vignette

is marked by an explanatory cartouche to ensure recognition. One of the more involved and

notable scenes occurs in the upper left of scroll two, just under the scene of Vulture Peak. This

large, brilliant vignette comes from the ―Chapter on the Dream of the Golden Drum of

Confession.‖96 The chapter opens with a vivid description of Myōdō‘s dream of a golden drum

which shone like the sun and sounded the beat calling forth confession. In his dream, a multitude

of Buddhas appeared, illustrated in the vignette encircling the golden drum as it emits light and

94
Hayashi makes this statement in regard to the atypical representation of the Four Guardian Kings in the mandalas.
See Hayashi, ―Daichōjuinzō konkōmyōsaishōōkyō kinji hōtō mandara zu oboegaki,‖ 88.
95
Jpn. Nyorai jūryō bon, Ch. Rulai shouliang pin; 如來寿量品. The scroll measures 140.2 cm x 54.6 cm. The
chapters represented in this scroll are the ―Introductory Chapter‖ and the second chapter, ―The Longevity of the
Tathāgata.‖
96
Jpn. Muken konku zange bon, Ch. Mengjian jingo chanhui pin; 夢見金鼓懺悔品. This scroll measures 140.2 cm x
54.3 cm and includes chapters are chapter three, ―Distinguishing between the Three Bodies,‖ and chapter four,
―Dream of the Golden Drum of Confession.‖

79
illuminates the world.97 Through the confession of sins, great rewards are promised. Once again,

the reoccurring theme of salvation and rewards through penitence is stressed.

Yiengpruksawan emphasizes the appropriateness of this overarching topic for the aging

Hidehira, and ten Grotenhuis even suggests that the crouched layman in red, crawling towards

the drum between the rays of light could be an image of Hidehira, perhaps in the guise of the

Brahman who beats the drum.98 Scroll three replaces the typical scene of the Vulture Peak

lecture in the upper right with an image of the Six Paths from the chapter, ―Elimination of

Karmic Obstructions.‖99 In this chapter, the proper avenue of repentance is further elaborated.

Both the scene and the cartouche identify this vignette as the moment when Śākyamuni enters a

state of deep concentration and illuminates the Six Paths of suffering from his follicles. 100 The

fourth scroll renders a bright sun burning above a fantastical carriage pulled by a white horse and

conveying a king surrounded on clouds by his entourage.101 No cartouche specifically identifies

this scene but Miya associates it with the highly visual description of the ten stages of bosatsu

wisdom (十地菩薩 jūji bosatsu) in the Golden Light Sūtra.102 He interprets the sun as the

ultimate stage of understanding, that of the Buddha, and the image of the king, identified as the

wheel-turning king, as the ninth stage of wisdom.

An interesting visualization of a passage occurs in the lower right of the fifth scroll, but

again lacks an explanatory cartouche.103 This vignette visualizes the metaphor of blowing the

97
Ishida draws a visual parallel between the drum depicted in the mandalas and the national treasure held in the
collection of Kōfukuji 興福寺, presumably the graceful drum supported by a lion with intertwined dragons (銅造華
原磬 dōzō kagenkei). See Ishida, ―Kokuhō saishōōkyō kyōtō mandara,‖ 7.
98
Yiengpruksawan, Hiraizumi, 200-01. Ten Grotenhuis‘ idea is from a personal communication published in Ibid.
99
Jpn. Metsugōshō bon, Ch. Mieyezhang pin; 滅業障品. The scroll measures 140 cm x 54.3 cm. This scroll presents
chapter five, ―Elimination of Karmic Obstructions.‖
100
T. no. 665, 16: 413c19-21.
101
This scroll measures 139.8 cm x 54.6 cm and offers chapter six, ―The Dhāraṇī of Absolute Purity.‖
102
Miya, Kinji hōtō mandara, 16. The excerpt from the sūtra comes from T. no. 665, 16: 419b27-c3.
103
This scroll measures 140.3 cm x 54.3 cm. The fifth scroll presents the chapter seven, ―Homage through the
Metaphor of the Lotus,‖ chapter eight, ―The Golden King Dhāraṇī,‖ chapter nine, ―Teaching on the True Nature of

80
conch of the great teaching 104 through the figure of a man exhaling into a conch shell atop the

right-most mountain.105 Below figures struggle in the water beside a boat ferrying others to

safety, again a visualization of another metaphor of salvation from the ocean of life and death. 106

The sixth scroll transcribes the twelfth chapter which is dedicated entirely to the nation

protecting Four Guardian Kings, and once again visually explains the proper methods for the

ritualistic veneration of the scripture.107 For instance, the central scene on the right side of the

mandala expresses the exact prescriptions to be undertaken for purified sūtra veneration, detailed

in both accompanying cartouche and graphic narrative. The scene describes a king, attended by

two figures, humbly holding an incense pot before a seated monk preaching the sūtra from within

a small hall. This moment references Tamonten‘s instructions in the sūtra to hold an incense

burner and make offerings, to cleanse the body and don clean robes, and in one quiet room

memorize and say the dharani. 108 The seventh scroll109 depicts the passage from the scripture

praising the prophylactic powers of the ‗Electric Kings‘ (電王 dennō) represented as two pairs of

Fūjin (wind god; 風神 Ch. Fengshen; Skt. Vāyu) and Raijin (thunder god; 雷神 Ch. Leigong; Skt.

Varun), symbolizing the four directions. 110 Underneath, a man uses a mallet without fear of

injury, and Miya further connects the images of the animal-headed deities as an elaboration of

Emptiness,‖ chapter ten, ―Reliance on Emptiness Fulfills Prayers,‖ and chapter eleven, ―The Four Guardian Kings
Observation of Humans and Deities.‖
104
T. no. 665, 16: 424c9.
105
Miya, Kinji hōtō mandara, 16.
106
T. no. 665, 16: 424c12. Miya, Kinji hōtō mandara, 16.
107
The sixth scroll measures 140.2 cm x 54.5 cm.This scroll contains the twelfth chapter, ―Protection of the Country
by the Four Guardian Kings.‖
108
T. no. 665, 16: 430c19-20.
109
The seventh scroll measures 140 cm x 54.3 cm and offers chapter thirteen, ―Dhāraṇī of Undefiled Attachment,‖
chapter fourteen, ―Wishing Fulfilling Jewel,‖ and chapter fifteen, ―Sarasvatī.‖
110
T. no. 665, 16: 433b16-21.

81
the passage, emphasizing the protective abilities of the deities who are able to keep the

animalistic deities at bay. 111

The upper left vignette in the central band of the eighth scroll pictorializes chapter

eighteen‘s passage concerning appropriate procedures for times when one falls ill.112 According

to the instructions, an afflicted person should retire to a room and cleanse the body. In the

illustration, an open-air room awaits its infirm occupant while Kichijōten kneels on a lotus

pedestal just outside. 113 The ninth scroll114 illustrates the mountainous retreat of a monk who

sought a simple existence in order to concentrate fully on the Golden Light Sūtra and so quit the

palace and its distractions. 115 The scene appears in the middle of the right side of the mandala

and shows the monk ardently venerating the sūtra, while a king, indicated by a parasol, bows in

worship as well. The tenth scroll116 features a rendition of a famous tale from the past lives of the

Buddha.117 In three scenes stacked along the right side of the mandala, the story of the hungry

tiger and the ultimate bodily sacrifice of the prince, a past reincarnation of Śākyamuni, is

depicted in continuous narration.118 In the first scene, the prince and his two brothers and father

enter the mountainous landscape. Below this, the princes explore the rugged landscape apart

from their father. In the final vignette, the climax of the tale is described. The prince is moved to

compassion for a starving tiger and her seven hungry cubs, and so prays and readies himself for

his sacrifice, depicted as the standing figure to the left of the action. The next moment renders
111
Miya, Kinji hōtō mandara, 18.
112
The eighth scroll measures 140.3 cm x 54.3 cm and offers a continuation of ―Sarasvatī‖ with the other chapters,
―Kichijōten,‖ ―Increase of Wealth by Kichijōten,‖ ―Pṛthivī,‖ ―Samjnaya, the Great General of the Yakṣas,‖ and
―Treatise on Correct Royal Rule.‖
113
Ishida, ―Kokuhō saishōōkyō kyōtō mandara,‖ 11
114
This scroll measures 140 cm x 54.3 cm. The ninth scroll includes the chapters: ―Susaṃbhava,‖ ―Protection by
Gods and Yakṣas,‖ ―Prophecy,‖ ―Pacification of Illness,‖ and ―Jala vāhana.‖
115
Ishida, ―Kokuhō saishōōkyō kyōtō mandara,‖ 12.
116
The tenth scroll measures 140.7 cm x 54.3 cm. This scroll offers the following chapters: ―Complete Giving of the
Body,‖ ―Praise by the Bodhisattvas of the Ten Directions,‖ ―Praise by Myōdō Bodhisattva,‖ ―Praise by the Goddess
of the Bodhi Tree,‖ ―Praise by Benzaiten,‖ and ―Entrustment.‖
117
T. no. 665, 16: 450c21.
118
T. no. 665, 16: 451a24-452b5.

82
the prince as he catapults himself off the cliff. And finally, the tiger and her cubs feed upon his

broken and bloodied body lying at the base of the precipice. 119

Of the three sets, the Chūsonji version is perhaps the most repetitive, a feature

characterized by Miya as formulaic and largely devoid of setsuwa-style images, resulting in a

composition which emphasizes overall balance and decoration. 120 The bottom of each Chūsonji

mandala shows scenes of gathering worshipers donning Chinese-style robes, 121 grouped and

scattered throughout gently rolling green hillocks and thin, wafting clouds of silver and gold.

Red bridges reach across the waterways—another theme common to Heian period

illustrations, 122 and one which Ishida Mosaku suggests represents the overcoming of disaster. 123

The overall effect of the scenes, despite the elevated hills, is one of flatness. The vignettes read

as stacked images of compressed space. In addition to the scenes of supplicants in low landscape

settings, in the uppermost right and left corners—save for the upper left of fascicle eight—

orthographic descriptions of Vulture Peak (鷲峯山 Jpn. Juhōsan, Ch. Jiufeng shan)124 cap each

mandala. Groups of seated Buddhas attended by bosatsu and/or figures dressed as monks appear

directly below the images of Vulture Peak, which provides the location for the Buddha‘s lecture,

on the right side of each fascicle and on the left sides of each fascicle except for numbers two,

119
For more information about the gift of the body, see Reiko Ohnuma, Head, Eyes, Flesh, and Blood: Giving Away
the Body in Indian Buddhist Literature, (New York: Columbia University Press, 2007). Of course, in each scroll
there are several more scenes that directly connect with the scriptural content; the images chosen here were selected
for their relevance to the sūtra and for their visual prominence.
120
Miya, Kinji hōtō mandara, 121.
121
Some have compared the clothing style of these assembled female worshipers dressed in Chinese-style garments
to those appearing in the 1157-58 handscroll, Annual Affairs (年中行事絵巻 Nenjū gyōji emaki), commissioned by
Go Shirakawa 後白河天皇 (1127-92), and therefore have proposed a parallel dating of the mid-twelfth century. See
Miya, Kinji hōtō mandara, 31-32; Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 264; Kameda, ―Jūbun
saishōōkyō jūkai hōtō mandara,‖ 68; and Sudō and Iwasa, Chūsonji to Mōtsūji, 151.
122
Kameda, ―Jūbun saishōōkyō jūkai hōtō mandara,‖ 68.
123
Ishida, ―Kokuhō saishōōkyō kyōtō mandara,‖ 7.
124
Jacqueline Stone notes that while Mt. Gŗdhrakūţa translates as ―vulture peak,‖ the terms Ryōjusen, from the
Chinese Ling-chiu-shan, and the term washi no yama, found in classical waka, both mean ―eagle peak.‖ See
Jacqueline I. Stone, Original Enlightenment and the Transformation of Medieval Japanese Buddhism (Honolulu:
University of Hawai‘i Press, 1999), 414 n40.

83
six, and eight. Moreover, the upper right and left groups of fascicles four, five, seven, nine, and

ten, and the upper right of fascicles six and eight are nearly identical: a centrally seated Buddha

offering the teaching mūdra, flanked by two monk-like figures and two bosatsu with colorful

trees in the background. Furthermore, several other groups and individual figures are also

repeated, such as the variously posed Four Guardian Kings discussed above; the pairings of Fūjin

and Raijin in numbers one, seven, and nine; the seated and standing Kichijōten in numbers six,

seven, and nine; and a lone kneeling bosatsu facing right in fascicles one, four, and five

(although here the bosatsu turns his head).

Dharma wheels proliferate in the Chūsonji mandala, suggesting significance beyond the

visual reference to the Buddha‘s law, as interpreted by Ishida Mosaku. 125 Perhaps the

preponderance of golden wheels is meant to conjure notions of authority and submission, to

suggest a hierarchy of divine rule injected into the earthly political realm. Perhaps the wheel-

turning king is graphically symbolized as the wheel around which worshipers gather. Other

scenes also evoke submission to authority, such as the many groupings of Buddhas flanked by

lower-ranking figures and the multiple images of Fūjin and Raijin indicating, 126 as promised in

the sūtra, the wheel-turning king‘s dominance over even natural phenomenon. The mandalas‘

visual references to the Hiraizumi terrain, the prominence of women, and the scenes of personal

repentance, anxiety, and blended faith argue for a localized reading of such theological rule, thus

investing the divinely awarded power in the hands of the Ōshū Fujiwara.

125
Ishida, ―Kokuhō saishōōkyō kyōtō mandara,‖ 7.
126
Ishida Mosaku believes that the deities of wind and thunder represent disaster. This interpretation also supports
the notion that proper devotion to the sūtra allows for the control of the weather and the prevention disaster. Ibid., 6.

84
Set History: Danzan Shrine Mandalas127

Unfortunately, far less is known about the Danzan Shrine128 and Ryūhonji mandalas than the

Chūsonji set; the context of the commission and production and the intended purpose for the

paintings are vague. Only a few inscriptions remain to cast low light on the shadowy history of

the mandalas, and secondary Japanese scholarship has focused primarily on the Chūsonji set.

However, piecing together the ephemeral clues with a visual comparison of contemporary works

gains some purchase in the pursuit of a historical account for the mandalas. While the mandalas

are not widely discussed in the literature, the Danzan Shrine versions are touted as the best

example of the jeweled-stūpa mandalas based on the Lotus Sūtra129 and even as one of the best

examples of the blue and gold transcription style. 130

A few clues emerge from the ink inscription on the boxes housing the Danzan Shrine

mandalas. At the time of the inscription, the mandalas were designated as ―Lotus Mandalas (法

華曼陀羅 hokke mandara),‖131 an earlier designation which has been adopted in Miya‘s work.132

As Miya points out, the term ‗lotus mandala‘ carries connotations unrelated to the Danzan Shrine

mandalas.133 By examining several medieval texts, he determines two broad categories of lotus

mandalas.134 The more schematically arranged lotus mandala associated with esoteric Buddhism

(密教 Jpn. mikkyō, Ch. mijiao) and often used in the Lotus Sūtra rites (法華経法 hokekyōhō)

127
For full images of each of the Danzan Shrine mandalas and details of each scroll‘s narrative vignettes, see Miya,
Kinji hōtō mandara, plates 74-144.
128
Danzan Shrine, originally a temple dedicated to Fujiwara Kamatari 藤原鎌足 (614-69), was built in the late
seventh century in Sakurai, Nara. For an introduction to the shrine, see Nara National Museum, ed., Danzan jinja no
meihō: yamato no kamigami to bijutsu 談山神社の名宝: 大和の神々と美術 (Nara: Nara National Museum, 2004).
129
Miya, Kinji hōtō mandara, 122.
130
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 263.
131
This title is inscribed along the exterior seam of the double doors. For a transcription of the inscription, see Miya,
Kinji hōtō mandara, 86 n1.
132
Ibid., 39.
133
Ibid., 39-42.
134
Ibid., 40-42.

85
frequently features Śākyamuni and Prabhūtaratna sitting side-by-side within a jeweled-stūpa

framed by an eight-petal lotus, a reference to the eleventh chapter of the Lotus Sūtra, the

―Apparition of the Jeweled-Stūpa.‖ An example of this type is found at Taisanji 太山寺 in

Hyōgo prefecture.135 The other category is the narrativization of the twenty-eight chapters of the

Lotus Sūtra (法華経二十八品大意絵 Hokekyō nijūhachi bon daiie, often shortened to 大意絵

daiie). However, if the entry is sufficiently ambiguous, as often they are, then it becomes

difficult to ascertain whether the ‗lotus mandala‘ in the passage refers to the esotericized version

or the transformation tableaux-type; certainty is possible only if the mandala is described

visually, or if the full categorical title is used for the paintings of the twenty-eight chapters.136

It is interesting to consider then the conceptual implications of this categorization of the

Danzan Shrine jeweled-stūpa mandalas as lotus mandalas. Such a categorization suggests that

the paintings were emphasized for their transformation tableaux characteristics, which

pictorialized the content of the Lotus Sūtra in the form of vignettes encircling the textual stūpa.

Certainly, the jewel-stūpa mandalas as a group embody the transformation of the sūtra‘s passages

into graphic narratives, in particular the mandalas focusing on the Lotus Sūtra, a scripture

celebrated for its descriptive content. Miya places the Danzan Shrine set into the category of

transformation tableaux-style lotus mandalas, explaining that they are actually an example of the

pictorialization of the twenty-eight chapters of the scripture.137 Undoubtedly, while the Danzan

Shrine and Ryūhonji sets feature graphic interpretations of the sūtra‘s didactic tales with the

accompanying chapter titles and passages as affixed cartouches—a style corresponding directly

with the transformation tableaux-type lotus mandalas—the narrative vignettes are only half of

135
For an image see Ibid., 40.
136
Ibid., 42.
137
Ibid.

86
the jeweled-stūpa mandalas‘ composition. Therefore, while they incorporate direct

characteristics of the transformation tableaux-style images, these mandalas are a far more

complicated visual and conceptual affair, as explored in chapters four and five, and thus

deserving of their own distinct, albeit small, categorization of painting: that of a jeweled-stūpa

mandala.

Production Date and Commission Context

Technique

With no primary texts recording the date, commission, or function of the Danzan Shrine

mandalas, the only recourse left in narrowing down approximate answers is a visual analysis of

the technique and style of the paintings and a possible link to the founding of a nearby temple.

One of the few scholars to work on this set of paintings, Miya proposes a late twelfth-century

date for the mandalas.138 By examining the technique used in communicating the Lotus Sūtra

content in the form of narrative vignettes, a few distinctive stylistic features, and a possible

reason for their commission, a mid- to late-twelfth-century production seems likely for the

Danzan Shrine mandalas.

As Miya describes it, the technique of pictorializing the tales of the sūtra and arranging

the narrative vignettes harkens back to an older style of setsuwa production. 139 Based on his

study of setsuwa pictures, Miya finds commonalities between the strong emphasis on story-

telling in the narrativized vignettes of the Danzan Shrine mandalas and the older style of setsuwa

depiction in which highly detailed scenes illustrating the scripture are favored. 140 Whereas earlier

setsuwa pictures tended toward the more comprehensively rendered and narrativized style in

138
Ibid., 81-5.
139
Ibid., 81.
140
Ibid.

87
which as much detail as possible was included in the scenes, later setsuwa images trended toward

simpler compositions without much emphasis on storytelling and where symbols were often used

to convey the main thematic messages of the scripture.141 Using the late Kamakura period

examples of Lotus Sūtra transformation tableaux-style mandalas from Honpōji 本法寺142 in

Toyama prefecture and from Honkōji 本興寺 in Shizuoka prefecture, Miya contrasts the strongly

narrative scenes of the Danzan Shrine set with these condensed vignettes that communicate their

narrative content in a much more distilled fashion. 143 He also points out that the continuous

narrative technique found in early setsuwa pictures is maintained in Danzan Shrine‘s jeweled-

stūpa mandalas.144

Furthermore, the extensive scriptural passages copied in the form of at times rather long

cartouches reproduce the style of early setsuwa picture‘s text and image relationships in which

copious cartouches record lengthy excerpts from the sūtra.145 Each scroll offers from a minimum

of thirteen to upwards of forty-six cartouches, with a total of two hundred and four for the entire

set.146 The biographical painting of Shōtoku Taishi 聖徳太子 (574-622) (聖徳太子絵伝 Shōtoku

taishi eden), dated to 1069 in Tokyo National Museum‘s collection of Hōryūji treasures,

provides/evinces a very similar treatment of the cartouches. Despite the many restorations of the

biographical painting, which introduce uncertainty regarding the original content of the

cartouches, it is clear from the size of the strips of paper originally attached to the painting

141
Ibid.
142
For a thorough discussion of these mandalas, see Haraguchi Shizuko 原口志津子, ―Toyama, Honpōjizō
‗Hokekyō mandara‘ no zuzō kaishaku to kanjinsō Jōshin 富山, 本法寺蔵『法華経曼茶羅』の図像解釈と勧進僧
浄信,‖ Kyōto bigaku bijutsushigaku 京都美学美術史学 3 (2004): 27-66.
143
Miya, Kinji hōtō mandara, 81.
144
Ibid.
145
Ibid.
146
Ibid., 122.

88
surface that the cartouches recorded lengthy passages from the scripture.147 The cartouches in the

Danzan Shrine mandalas served an important function. The crowded composition of each

painting might make identification of the numerous scenes difficult, so the mandalas‘ many and

extensive cartouches clarify the connection of the graphic narrative to the written sūtra. Aside

from slight mistakes in copying and omissions, the cartouches remain faithful to the sūtra. Miya

intriguingly observes that if one was to rearrange the passages on each mandala into the correct

textual sequence, a form of the Lotus Sūtra emerges, albeit, abbreviated.148 In this way, the

mandalas present the scripture in its entirety in the form of the central textual reliquary, but also,

in the form of a truncated text occurring through the ten paintings. And given the difficulty of

reading the unabridged Lotus Sūtra manipulated into the shape of a stūpa, the cartouches

preserve the option of reading the scripture. Kamakura period setsuwa pictures often only record

the title and a few explanatory characters, such as the name of the deity depicted. Ryūhonji‘s

jeweled-stūpa mandala set follows this later style of text and image correlations; most of the

cartouches list only the title associated with the vignette with little copying of sūtra passages.

Danzan Shrine‘s mandalas offer condensed transcriptions of the Lotus Sūtra beside the

pictorialized versions of these passages, and in this way, reflect the relationship between text and

image in narrative handscrolls (絵巻物 emakimono). But rather than alternate in a clear,

consistent fashion where the usually extensive text known as kotobagaki 詞書 precedes a

painting made for a limited audience as is the case with many handscrolls, the mandalas must

utilize one continuous picture surface where contiguous vignettes are often unrelated and the

explanatory text and image are viewed simultaneously.

The cramped composition of the Danzan Shrine mandalas also harkens back to the high

147
Ibid., 80.
148
Ibid., 79.

89
surface density of earlier Chinese and Korean setsuwa pictures. This set offers the most narrative

translations of the Lotus Sūtra’s parables through approximately two hundred and thirty

illustrations, 149 creating a somewhat cramped composition with all available space devoted to

graphic description. The most beloved of the sūtra‘s tales are awarded more scenes so as to

capture fully the key moments of the story. For example, the popular parables of chapter twenty-

five, ―The Univeral Gateway of the Bodhisattva He Who Observes the Sounds of the World,‖

occupy the entire right column and a quarter of the left column of the composition in fascicle

eight with depictions of the perilous circumstances within the miraculous reach of Kannon

bosatsu‘s intercession. 150 Hamada Takashi compares the saturated picture surface of the

mandalas with the style of Song dynasty setsuwa-style images. 151 He posits that the jeweled-

stūpa mandala set at Danzan Shrine has a more direct connection with Chinese visuals than the

set at Chūsonji which embodies a more Japanese style, expressed through the arrangement of the

vignettes, the composition density, and the delicate Fujiwara painting style. 152 The techniques of

dividing the surface by mists and thereby segregating the vignettes, and the relatively consistent

picture configuration discussed under the narrative vignette section, are derived from earlier

Chinese and Korean compositions. 153 Overall, the visual techniques implemented in the Danzan

Shrine mandalas suggest an affinity to Chinese and Korean setsuwa-style pictures and compel a

late twelfth-century dating, which is supported by an analysis of the narrative vignette style.

149
Ibid., 74.
150
Jpn. Kanzeon bosatsu fumon bon; Ch. Guanshiyin pusa pumen pin; 觀世音菩薩普門品. This scroll measures
134.1 cm x 52.6 cm. and includes chapter twenty-five, ―The Universal Gateway of the Bodhisattva He Who
Observes the Sounds of the World,‖ chapter twenty-seven, ―The Former Affairs of the King of Fine Adornment,‖
and chapter twenty-eight, ―The Encouragements of the Bodhisattva Universally Worthy.‖ Chapter twenty-six,
―Dhāraṇi,‖ is not represented.
151
Hamada, ―Konkōmyōsaishōōkyō kinji hōtō mandara zu,‖ 263.
152
Ibid.
153
Miya, Kinji hōtō mandara, 122.

90
Style

Obvious connections can be drawn between the Danzan Shrine mandalas and the conventional

blue and gold illuminated sūtras. As has already been noted in the previous section, dating

frontispiece illustrations by style alone is a tricky task because of the relative uniformity of the

images across a broad span of time. However, based on the general characteristics of

frontispieces, early twelfth-century examples display a sensitivity of line and less patternization

of the composition. For instance, the frontispieces of the early twelfth-century Kiyohira sūtras

exhibit a delicateness of line with compositions full of fine-line detail. These earlier

compositions brimming with drawings that occupy most of the picture surface later trend toward

simpler compositions in the Kamakura period. The line quality of thirteenth-century frontispieces

generally is harder and stiffer and the compositions more formatted and standardized.

Much like the busy compositions of the twelfth century, the Danzan Shrine mandalas fill

all available space with sensitively rendered fine-line detail. There is a hardening of line,

particularly evident in the rock and cliff faces which appear more formatted and drawn with

firmer lines that lack gentle modulation. The figures of the deities and people along with the

flames of fire and waves of water are drawn with lines thin and lively, reminiscent of twelfth-

century blue and gold sūtra frontispieces. But overall, the lines appear stiffer than earlier

examples, suggesting a date closer to the Kamakura period than the mid-Heian.

As discussed above, the influence of Song dynasty stylistic characteristics is evident in

the Danzan Shrine mandalas. Several stylistic choices made in the mandalas do not conform to

the general characteristics of blue and gold illuminated frontispieces. Bodies of brilliant gold

paint for both deities and people, red paint for the pinched mouths, black ink for the eyes, noses,

and fingers, and white paint for hair are all uncommon visuals for Heian period frontispieces; yet

91
a miniature Lotus Sūtra blue and gold illuminated scroll in the possession of Danzan Shrine

shares these distinctive stylistic qualities. Likely of Korean provenance, this seventh volume of

the Lotus Sūtra was probably an eleventh-century product.154 The frontispiece for this volume

depicts the familiar scene of Śākyamuni preaching to a gathered crowd. The bodies of the

assembled deities shine forth in gold while their lips are rendered red and details of the body

delicately drawn using black ink. The meticulous drawing expands to fill the composition on all

sides. The lines are elegant and slender, and, though the format of the scroll is a miniature, the

sensitivity to detail is finely executed. Certainly, the impact of Song style compositions and

linework reaching Japan in the form of such images as sūtra copies and setsuwa-style pictures of

both Chinese and Korean origin can be seen in many remaining examples of the tenth through

thirteenth centuries. Another example might have influenced the production of the Danzan

Shrine mandalas: the blue and gold ichiji hōtō hokekyō (one character, one stūpa sūtra) of 1163

attributed to the monk Shinsai. The frontispiece of this Lotus Sūtra copy also exhibits

characteristics of Song dynasty style sūtra paintings. According to the vow, Shinsai produced

this project in hopes of entering Amida‘s paradise. 155 Miya believes that this scroll could have

been an original source for the mandalas. 156 Pinpointing an exact and single source for the

Danzan Shrine jeweled-stūpa mandalas is difficult and perhaps even misguided, as a variety of

sources were used in making the mandalas. The mandalas display characteristics taken from

paintings ranging from Japanese twelfth-century blue and gold illuminated sūtras—the impact of

which is seen most prominently in the choice of the blue and gold material and the drawing

techniques of the landscape—to Chinese and Korean sūtra and setsuwa examples, demonstrated

through the choice of figure rendition and compositional techniques.

154
Ibid., 84. For an image of the scroll, see Ibid., 84 fig. 57.
155
Nara National Museum, Seichi Ninpō (Ninpō), 299.
156
Miya, Kinji hōtō mandara, 88 n60.

92
A Potential Commission

A tantalizing inscription from the outer lid of the box containing the mandalas ambiguously

mentions a temple roughly half a kilometer northwest of Danzan Shrine called Shigaiji. 157

Precious little is known about this elusive temple around the time of the mandalas‘ production.

The temple records of Danzan Shrine rarely refer to Shigaiji and when the temple does appear in

the literature, it is only in records far closer to present day than the twelfth century. 158 This

mortuary temple was founded in 1187 in honor of the Tendai monk, Zōga 増賀 (917-1003).159

Zōga‘s devotion to the Lotus Sūtra was renowned, and it is possible that this jeweled-stūpa

mandala set was commissioned for the founding of the temple to memorialize his dedication to

the scripture. As explored in the above sections, both the techniques and style of the mandalas

confirm a late twelfth-century production date. Furthermore, given the exterior inscription on the

box housing the mandalas, it seems likely that Shigaiji wished to commission a project worthy of

a temple founding and one with connotations specific to the establishment of the temple. Shigaiji

maintained the mandala set until the Meiji period (1868-1912). The founding of the temple in

1187 supports the date already concluded by the visual analysis of the mandalas and gives a

special function for the paintings. This possibility adds a commemorative function to the

jeweled-stūpa mandalas and stresses the transference of merit through the copying of the sūtra,

the adorning of the body of the Buddha with precious materials, and the construction of stūpas—

a karmic confluence particular to the commission of this rare type of project.

While the dating and commission status of Danzan Shrine‘s jeweled-stūpa mandalas has

no conclusive answer, a sound approximation is during the late twelfth century. The techniques

of highly detailed vignettes with accompanying long scriptural passages and an overall saturated

157
See Miya, Kinji hōtō mandara, 86 n1 for the inscription.
158
Ibid., 85.
159
Ibid.

93
composition employing mists as dividers and reminiscent of earlier Chinese and Korean

examples, the stylistic choices reflecting the impact of twelfth-century blue and gold illuminated

sūtras and Song dynastic styles, and the inscription prominently bearing the name of Shigaiji—a

temple founded for a monk devoted to the Lotus Sūtra—all come together to support a

production date of ca. 1187 and a commemorative commission for the Danzan Shrine mandalas.

Formal Analysis

Process of Construction

The jeweled-stūpa mandalas of Danzan Shrine consist of ten large paintings, a full transcription

of the twenty-eight chapters of the Lotus Sūtra in eight volumes with the bracketing scriptures,

Sūtra of Innumerable Meanings and Sūtra of Meditation on the Bodhisattva Universal Worthy.

Unfortunately, the mandalas of Danzan Shrine are not as well-preserved as the Chūsonji and

Ryūhonji sets. Faded scenes and damage to the surface of the paintings hinder the reading of

many scenes. Each scroll suffers from extensive creases, the result of having been stored in

rolled form for a long time. This is particularly true in the case of the Sūtra of Innumerable

Meanings and the Sūtra of Meditation on the Bodhisattva Universal Worthy, the opening and

closing fascicles respectively, where much of the textual stūpa has been worn away and a

majority of the surrounding scenes are almost illegible. The fine-line description of the narrative

vignettes and crowded composition further undercut the decipherability of the mandalas. Despite

these obstacles, because of their adherence to an older, continentally informed model, the

Danzan Shrine mandalas offer a rare and more individualized view of jeweled-stūpa mandala

composition.

The techniques required for the production of the Danzan Shrine set are much the same

94
as those employed for the creation of the Chūsonji mandalas, as well as the Ryūhonji mandalas.

Once again the scale of such an elaborate project necessitates a carefully executed pattern

mapping out the exact position of each story, each eave, and each architectural detail and how

each transcribed character from the sūtra would be manipulated to build the textual reliquaries.

Visible upon close scrutiny, tiny, indented lines running both horizontally and vertically lay out

the precise placement of each row of characters. Such involved pre-planning resulted in a

uniform and orderly presentation of golden manifestations of the malleability of sūtra text, at

once building a complex, textual stūpa and offering graphic narratives that surround the dharma

reliquary. The scrolls are composed of at least seven sheets dyed a deep indigo. 160 Two long

sheets joined directly outside the fourth story of the stūpa make up the right and left narrative

bands of the mandalas. Two sheets, again meeting just above the fourth story, form the central

band of the paintings, upon which is rendered the textual reliquary. The sheet attached under the

stūpa, usually after the ornithologic representation of Vulture Peak and within the upper halo

section of Śākyamuni, contains the bottom narrative descriptions. The joints are disguised by

landscape and architectural details that obscure the horizontal merger of the papers, but the

distinction of the sheets forming the right and left bands from the central stūpa is marked. A

small, though nonetheless visible vertical line runs the length of the mandalas at the intersection

of the outer bands with the central band. Perhaps because of the wear and age of the mandalas,

the thin, overlapping strips of paper where the glue holds the sheets together has become very

noticeable and the paint atop it has faded dramatically in places. Much like the already described

Chūsonji procedures, the sheets used to fashion the right and left narrative bands are often dyed

160
Even after having personally viewed the mandalas in storage at Nara National Museum in January 15, 2009, it is
difficult to determine whether the lower right and left corners are small, separate sheets. I am very grateful to
Donohashi Akio for arranging the visit to the museum‘s storage and for the extensive photographs he took of the
paintings.

95
the same hue. The central sheets were dyed a matching hue, but the extant difference in color

between the middle two sheets on which the stūpa is rendered on the one hand, and the outer four

sheets which host the narrative vignettes on the other, suggests an exacting order to this vast

project, much like an artistic assembly line production. Left virtually empty apart from the

description of waves and the occasional entrance of a vignette in the Chūsonji and Ryūhonji

versions, the space around the stūpa is compressed in the Danzan Shrine mandalas. The graphic

interlopers leave little negative space, often filling in the areas between the eaves. Because of the

difficulty in gauging the placement of the vignettes that exceed the boundaries of the bordering

bands without the stūpa already constructed, it seems highly likely that the stūpa was completed

first and then the sheets already containing the vignettes were added and the narrative sections

expanding out into the realm of the stūpa were finished at the end. A thin one centimeter

arabesque band frames the sides of the mandalas while thicker bands of arabesque cap the top

and bottom.

Although the circumstances of the Danzan Shrine mandala production remain elusive,

more than one brush can be identified at work in the illustrations and thus it is improbable that a

sole artist painted the narrative vignettes. As artists were likely specialized, copyists trained in

sūtra transcription were probably commissioned to produce the central textual stūpa while artists

specializing in painting depicted the plentiful narratives. Interestingly, the mandalas have a slight

green hue, attributable to the higher concentration of copper in the gold.161

Stūpa

A consistent feature of the jeweled-stūpa mandalas is the careful attention to architectural

accuracy. While the Danzan Shrine mandalas do not feature stairs leading up to the stūpa like the
161
Miya, Kinji hōtō mandara, 87 n26. Miya also describes the muddied gold as ―blue gold.‖ See Ibid., 119.

96
Chūsonji and Ryūhonji versions, they do incorporate the main components of stūpa structures.

Unfortunately, the finials of the mandalas have suffered extensive wear, making it difficult to

appreciate the full crowning apparatus for its accuracy and delicate construction. Each do include

the seven appropriate attributes already described in the Chūsonji section: (1) the precious jewel

at the top, (2) the oval element between the jewel and (3) the metal decoration which is affixed to

the exposed central pillar, (4) the nine metal rings of the central pillar, (5) the simple flower base

upon which the above structure rests, (6) the inverted bowl which supports the flower base, and

(7) the square box hosting the full finial. Again, base rafters, bracket arms connecting the beams

to the pillars, and the bearing blocks for the bracket structure make up each floor of the stūpa.

The stacked windows on each story issue forth golden light, suggesting a brilliant realm inside

the stūpa. The foundation of the reliquary rests upon the common motif of an inverted lotus base,

drawn with regular line rather than scriptural text. This version foregoes the banister framing the

foundation. Unlike the other sets, the Danzan Shrine mandalas feature a foundational floor made

of sūtra characters and spreading out before the single Buddha with a bright red ūrṇā (the tuft of

white hair between the Buddha‘s eyebrows that issues forth brilliant light revealing all the

worlds; 白毫 Jpn. byakugō, Ch. baihao). Oxidized panels of silver flank the doors composed

entirely of sūtra text, which open up revealing the seated Buddha who offers the preaching

mūdra. While the stūpa at first glance appears to be a ten-story structure, the first roof is actually

a false roof, making it a nine-layered stūpa, much like the architecture in the Chūsonji mandalas.

The graceful, undulating waves in the background of the textual reliquary draw out the stūpa and

focus the eye upon it.

The most distinctive feature of the jeweled-stūpa mandala format is surely the sūtra text

97
manipulated into the shape of a stūpa. Shallow grooves made by a metal stylus162 forge sharp

lines that create balanced rows and columns of diminutive characters to fashion the textual

reliquary. Like the pattern of jeweled-stūpa mandalas, the transcription begins at the top most

detail of the finial: the tip of the crowning jewel. From this height, the sūtra continues in order,

flowing down the stūpa and filling out the nine stories. Each volume is precisely calculated to

finish with the last words of the last chapter in that volume, followed by the title of the sūtra and

the volume number. This sequence forms the last few characters of the bottom line of the stūpa,

finishing in each mandala with the lower-most right character of the foundation. 163 Unlike the

Chūsonji mandalas which occasionally need to repeat portions of the volume in order to reach

the concluding point, the Danzan Shrine mandalas are accurately measured to close in this very

specific technique.

Narrative Vignettes

Miya records the narrative vignettes of each scroll in exacting detail, so I will not repeat this

catalog.164 Instead, this section summarizes the main compositional themes and allocation

choices to emphasize the decisions faced by the artists when constructing such a complicated

structure. Unlike the tidily distinct vignettes of Ryūhonji and in particular Chūsonji, the scenes

surrounding the central stūpa in the Danzan Shrine mandalas seem a cohesive narrative mass,

engendering more than a little confusion as scenes partake of the same shared space but without

a connection in content. The density of the compositions results from the sheer number of

narrative vignettes and the extensive use of mountainous scene partitions. Further accentuating

the sense of surface saturation, these many scenes are firmly grounded in golden, hilly

162
Ibid., 42.
163
Ibid., 119.
164
See Ibid., 44-76. For a discussion of Miya‘s visual analysis, see the literature review in chapter one.

98
landscapes populated by numerous figures associated with the narratives. In fact, the Danzan

Shrine version offers by far the most landscape settings of the three sets, nearly filling all

available space. The gradually sloping hills and rocky mountain faces have golden washes as a

base with thin lines of gold—both parallel and cross-hatched—on top to add descriptive

landscape marks and volumetric development. And while scale is certainly not a priority, this set

is more dimensionally faithful, achieved in part because the mountains are much higher and

steep than the landscape depicted in the other two sets. Within each scene, the profusion of

delicate, fine-line detail heightens the sense of graphic density and offers full descriptions of

even the minutest components, giving the composition a unique richness since vitiated by

extensive fading and wear. The compositions are also much more populated in the Danzan

Shrine mandalas when compared with those of the Chūsonji and Ryūhonji sets. Each fascicle,

transcribing one volume of the Lotus Sūtra composed of two to six chapters in the main eight

scrolls, depicts between six to forty-five scenes165 with an astonishing total of two hundred and

thirty vignettes.166

The arrangement of the narrative vignettes can be generalized, with some exceptions, as a

relatively consistent circular assembly around the textual stūpa. In particular, the allotment of

graphic tales are broadly grouped by chapters around the reliquary but within the space assigned

to a particular chapter, the arrangement of the scenes is often random, disobeying the

chronological rules of the scripture. The general rule is that the scenes begin in the lower left

under the stūpa, flowing upward to the top of the narrative band before crossing over to the right

side and continuing downward to the concluding scene at the bottom of the right of the stūpa.

The narrative path begins with the consistently repeated scene of Vulture Peak at the base of the

165
Ibid., 44.
166
Ibid., 74-75.

99
stūpa; from this point, the route of the vignettes assumes the shape of a calligraphic hiragana no

の.167 He explains that this system of allocation conforming to the circular no pattern originates

in the wall paintings of Lotus Sūtra transformation tableaux at Dunhuang (敦煌 Jpn. Tōnkō).168

The bottom-up method of allocation can also be seen in examples from central India and central

Asia.169 From these observations, he concludes that the Danzan Shrine mandalas follow the

conventional arrangement of narratives associated with Buddhist setsuwa pictures,170

consequently conveying and preserving a more continental sense than the other sets.171 The

system of vignette arrangement differs greatly between the three sets, indicating another facet in

which the jeweled-stūpa mandalas as rare paintings existing on the periphery of artistic

production skirt the standardization often seen in other categories of paintings. Partly this is

attributable to the flexible nature of transformation tableaux and setsuwa picture organization—a

visual stream of which the jeweled-stūpa mandalas partake172—but also because the rarity of this

format allows for great flexibility, especially given that the production of these paintings were

occurring at vastly different places for singular spaces. Clear exceptions to this rule are the first,

fourth, and eighth scrolls. The first scroll, 173 perhaps because it only holds two chapters, is

divided roughly in half along the vertical axis with chapter one occupying the lower half and

chapter two assuming the upper half. Rather than start in the lower left, the fourth fascicle 174 has

167
Ibid., 76,120.
168
Ibid., 120.
169
Ibid.
170
Ibid.
171
Ibid., 148.
172
Miya classifies the mandalas as transformation tableaux, and in particular, the Danzan Shrine set as Lotus Sūtra
transformation tableaux. Ibid., 42.
173
The first fascicle measures 133.3 cm x 52.8 cm, and includes the first chapter, ―Introduction,‖ and the second
chapter, ―Expedient Devices.‖ See Ibid., 74-76 for a clear list of the scenes depicted in each mandala.
174
The fourth fascicle measures 134.3 cm x 52.6 cm. This scroll contains the fourth volume of the Lotus Sūtra,
including the chapters: ―Receipt of Prophecy by Five Hundred Disciples‖ as chapter eight, ―Preachers of the Dharma‖
as chapter ten, and ―Apparition of the Jeweled-Stūpa‖ as chapter eleven. Chapter Nine, ―Prophecies Conferred on
Learners and Adepts,‖ is not included.

100
the minor alteration of beginning in the lower right. However, the eighth scroll breaks the

convention almost entirely. Because of the popularity of the twenty-fifth chapter which

champions the intercessory might of Kannon, the majority of the composition is dedicated to

representations of its themes, perhaps to the detriment of the twenty-sixth chapter which is

completely elided. The whole of the right narrative band depicts miraculous scenes from the

twenty-fifth chapter, while the left narrative band is quartered with two sections belonging to

chapter twenty-eight, another quarter devoted to chapters twenty-five, and one section

representing chapter twenty-seven‘s episodes.

Another spatial rule governing the placement of the narrative episodes are the themes of

paradise and hell. Vignettes depicting heavenly scenes are typically grouped in the upper

registers of the mandala, often jettisoning the chronological arrangement within their chapter

allocation around the stūpa. In the first scroll, a paradisiacal scene of Akaniṣṭha Heaven (阿迦尼

吒天 Jpn. Akanita ten, Ch. Ajianizha tian) breaks from its fellow scenes illustrating the first

chapter, located in the bottom half of the scroll, in order to be thematically representative. The

scene of floating palatial architecture is identified with a short caption taken from the sūtra. 175

Appropriately, the celebrated apparition of Prabhūtaratna‘s stūpa in the air above the gathered

crowd at the time of Śākyamuni‘s lecture of the Lotus Sūtra is described in a large scene on the

upper right of the fourth scroll. 176 In an example of conflated space, a heavenly palace drawn in

the upper left corner of the eighth scroll serves as both an illustration of Trāyastriṃśa Heaven (忉

利天 Jpn. Tōri ten, Ch. Daoli tian) and of Tuṣita Heaven (兜率天 Jpn. Tosotsu ten, Ch. Doushuo

175
The caption describes reaching up to Akaniṣṭha Heaven. T. no. 262, 9: 2b18. See Miya, Kinji hōtō mandara, 45
for a diagram explaining the location of the chapters and accompanying cartouches on the first mandala.
176
The passage describes the thunderous moment when Śākyamuni opens the door of the recently manifested stūpa
with his right finger. Leon Hurvitz, trans., Scripture of the Lotus Blossom of the Fine Dharma (New York: Columbia
University Press, 2009), 167. T. no. 262, 9: 33b26-28. Miya notes that the first two characters have been reversed.
See Miya, Kinji hōtō mandara, 56. For the diagram of the fourth mandalas showing the arrangement of the chapters‘
scenes and cartouches, see Ibid., 55.

101
tian). Without the explanatory cartouches, this dual representation would be difficult to

identify. 177 And likewise, the explicit description of the torture and misery in the lower paths of

existence populate the lower registers of scrolls one and six. The bottom corners of scroll one

stage grotesque scenes of violence and suffering: scenes of Avīci Hell (阿鼻地獄 Jpn. Abi jigoku,

Ch. Abi diyu) act as the complement to the topmost scene of Akaniṣṭha Heaven, emphasizing the

expansive and limitless reach of the Buddha‘s light, a direct connection to the scene under the

stūpa in which the golden rays reach even hell. 178 On the right side above the scene of warring

gods (阿修羅 Jpn. ashura, Ch. axiuluo), hungry ghosts dwell in anguish, the water they attempt

to drink turning into flames and their necks—as thin as pins—forbidding the quenching passage

of sustenance. Further scenes of hell are illustrated in their thematically accurate position in the

lower right corner of the sixth scroll. 179 The cartouche to the right of the stūpa‘s foundation

records the torturous screams issuing from hell and the cries of the hungry ghosts as they

desperately search for food and drink. 180 This quote is continued in the lowest cartouche on the

right side narrative band in which the terrifying voices of the warring gods living by the edge of

a vast sea are described. 181

Other circumstances also oblige bending the spatial rules. For instance, interloping
177
The longer excerpt directly under the palace explains specifically praises the transcription of the Lotus Sūtra,
promising a favorable rebirth in Trāyastriṃśa Heaven to the glorious music of eighty-four thousand female deities.
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 307. T. no. 262, 9: 61c4-6. Miya explains that the
passage detailing the eighty-four thousand female deities is omitted. See Miya, Kinji hōtō mandara, 70. A cartouche
also identifies the palace as Miroku‘s Tuṣita. Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 307. T. no.
262, 9: 61c10. For a diagram of the location of the chapter scenes and cartouches, see Miya, Kinji hōtō mandara, 67.
178
The cartouche identifies this scene as the lowest realm of hell, Avīci. Hurvitz, Scripture of the Lotus Blossom of
the Fine Dharma, 5. T. no. 262, 9: 2b17-18. Another explanatory cartouche just outside the wall of hell reiterates in
verse (偈他 Jpn. geta; Ch. jieta; Skt. gāthā) the Buddha‘s salvific reach throughout the six realms. Hurvitz,
Scripture of the Lotus Blossom of the Fine Dharma, 7. T. no. 262, 9: 2c16-17.
179
This scroll measures 134.5 cm x 52.6 cm, and transcribes the sixth volume‘s ―The Life Span of the Thus Come
One‖ as chapter sixteen, ―Discrimination of Merits‖ as chapter seventeen, ―The Merits of Appropriate Joy‖ as
chapter eighteen, and ―The Merits of the Dharma Preacher‖ as chapter nineteen.
180
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 244. T. no. 262, 9: 48a19-20. For Miya‘s diagram of
the sixth scroll detailing the placement of the chapter scenes and cartouches, see Miya, Kinji hōtō mandara, 60.
181
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 244. T. no. 262, 9: 48a21-22. See Miya for the
minor alterations made to the passage as quoted in the mandala. Miya, Kinji hōtō mandara, 62.

102
chapters enter the area allocated to another chapter, and the resulting merger creates a space

ordered for compositional rather than chronological convenience. As a continuous single picture

surface, space is at a premium and scenes are grouped together despite being from different

chapters. In the upper right register of scroll seven, 182 narratives from chapters twenty-two,

twenty-three, and twenty-four all occupy the same space because, such visual mergers take

advantage of similar landscape, identical main characters, and other factors determining the

composition of the vignettes. Such examples indicate that one graphic space can hold multiple

meanings. In this case, the narratives describing the actions of Bodhisattva Fine Sound (妙音菩

薩 Jpn. Myōōn bosatsu, Ch. Miaoyin pusa; Skt. Gadgadasvara), for whom the twenty-fourth

chapter is named, and the Bodhisattva Seen with Joy by All Living Beings (一切衆生憙見菩薩

Jpn. Issai shūjō kiken bosatsu, Ch. Yiqie zhongsheng xijian pusa; Skt. Sarvasattvapriyadarśana)

take place in the heavens and share the thematic connection of praising the Buddhas, and

therefore supply the reasons for the amalgamation. The scene along the right edge conveys the

story of Myōōn bosatsu‘s offerings to the Buddhas.183 Also from the same chapter directly below

this scene are a vignette and cartouche describing the arrival of Myōōn bosatsu with a retinue of

eighty-four thousand bosatsu in order to further praise the Buddha Pure and Bright Excellence of

Sun and Moon (日月淨明德 Jpn. Nichigetsu jōmyō toku, Ch. Riyue jingming de; Skt.

Candrasūryavimalaprabhāśrī). 184 The scene along the finial describes Bodhisattva Seen with Joy

182
Scroll seven measures 133.6 cm x 52.8 cm. The chapters traditionally assigned to volume seven are all included:
chapter twenty, ―The Bodhisattva Never Disparaging,‖ chapter twenty one, ―The Supernatural Powers of the Thus
Come One,‖ chapter twenty-two, ―Entrustment,‖ chapter twenty-three, ―The Former Affairs of the Bodhisattva
Medicine King,‖ and chapter twenty-four, ―The Bodhisattva Fine Sound.‖ For the diagram of the chapter scenes and
cartouches, see Miya, Kinji hōtō mandara, 63.
183
Given the convergence of several chronologically unrelated scenes, cartouches serve an important identification
function. The cartouche for this scene explains the all-penetrating divine light issuing from Śākyamuni‘s ūrṇā.
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 277. T. no. 262, 9: 55a17-18.
184
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 279. T. no. 262, 9: 55c7-9.

103
by All Living Beings‘s worship of the Buddha and provides the associated verses. 185

As discussed in the Chūsonji mandalas, an inherent characteristic of transformation

tableaux-style imagery is the technique of continuous narration. Given the Lotus Sūtra‘s strong

proclivity toward narrative explanation, there are numerous examples within the Danzan Shrine

mandalas of narratives depicting the passage of time within the same vignette. Miya‘s exhaustive

examination of the vignettes lists the tales utilizing this visual technique, 186 such as the Burning

House Parable from the third chapter, illustrated in the second scroll. 187 In this series of related

vignettes, the story begins at the bottom of the left narrative band and continues upward,

culminating in the salvation of the great man‘s three sons. This narrative flow matches the

general motion of the vignettes around the textual stūpa. The visual technique employed here can

also be seen in handscroll illustration. However, unlike the unfurling of a handscroll in which the

scenes follow one another and directly relate, the challenges of structuring usually disparate

scenes without any interruptions to the picture surface, for instance the interspersions of narrative

text, forces the viewer to act as a visual detective who sequences the depictions, in effect

transforming the viewer into a narrator of the sūtra.

At the bottom center of each Danzan Shrine mandala is a large gathering of figures

around Śākyamuni who, seated on a raised dais on Vulture Peak, delivers in perpetuity the very

sermon depicted above him in the form of a reliquary. The geography of Vulture Peak is also

symbolized by an ornithologic hill in each fascicle. 188 While differences exist in who is

represented as well as the numbers of figures present, the most common formula for the

185
The cartouche offers the verses praising the Buddha‘s fine countenance and brilliant light which penetrates all ten
quarters reads. Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 272. T. no. 262, 9: 53c4-5.
186
For his full list, see Miya, Kinji hōtō mandara, 77.
187
This scroll measures 133.5 cm x 52.8 cm, and includes the third chapter, ―Parable,‖ and the fourth chapter,
―Belief and Understanding.‖
188
Because I was unable to view the tenth fascicle featuring the closing sūtra and because the reproductions of this
scroll in Miya‘s book and in exhibition catalogues are unclear, I am unable to say whether this is the case for scroll
ten.

104
gathering features Śākyamuni offering the teaching mūdra surrounded by Monju on his right,

Fugen on his left (identifiable by their respective animal vehicles: the lion and the elephant),

sixteen rakan, and eight to ten guardians. In comparison to the richly and rigorously repetitive

mandala of Chūsonji, the Danzan Shrine scrolls are much more varied.

Set History: Ryūhonji mandalas189

Much like the Danzan Shrine version, mystery shrouds the patronage of the Ryūhonji jeweled-

stūpa mandalas. And like the Danzan Shrine set, an alluring yet elusive inscription hints toward a

complicated historical journey for the Ryūhonji mandalas. On the back of each scroll, a black ink

inscription records the mandalas‘ location in Hōryūji at the time of its first recorded restoration

in the seventh month of 1362.190 Whether Hōryūji was the original home of the paintings or just

the first recorded residence is unclear. However, it is clear that by the mid-fourteenth century the

mandalas were in the possession of Hōryūji. A later inscription testifies to another restoration in

1681 in Edo (modern day Tokyo).191 At this time, the mandalas received new mounting as a

gift.192 In volume nineteen of Taishiden gyokurin shō 太子伝玉林抄, a list of Hōryūji‘s treasures

documents eight Lotus Sūtra stūpas (法花八塔 Hokke no hattō) and a 1483 inventory also

records eight Lotus Sūtra stūpas housed in a box. 193 Miya concludes that these perfunctory

entries refer in fact to the jeweled-stūpa mandalas.194 According to Ogino Minahiko 荻野三七彦,

these writings are records from the inspections of the temple treasures known as the Shariden

189
For full images of each of the Ryūhonji mandalas and details of each scroll‘s narrative vignettes, see Miya, Kinji
hōtō mandara, plates 145-205.
190
Miya, Kinji hōtō mandara, 90. Miya provides a transcription of the inscription along with photographic evidence
of the inscription from the back of the first scroll.
191
Ibid.
192
Ibid.
193
Ibid.
194
Ibid.

105
hōmotsu 舎利殿宝物.195 According to Ogino‘s research, there were four inspections made into

the collection which included the mandalas, the first one was made in 1550 and is the entry in the

Taishiden gyokurin shō.196 The second inspection was conducted in 1591, and the third and

fourth inspections (which resulted in no records about the mandalas) occurred in 1609 and 1652,

respectively. 197 Based on these findings, Hōryūji was in possession of the mandalas until the

mid-sixteenth century, at which point they were transferred to Ryūhonji by the time of the 1681

inscription.

Production Date

Technique

Because little can be surmised about the commission context of the Ryūhonji jeweled-stūpa

mandalas, an approximate production date must derive from a visual analysis of the technique

and style of the paintings—much like the approach undertaken with the other two sets, although

here, contextual circumstances are unavailable to support the likely date or provide a glimpse at

any intended function. A cursory glance to compare and contrast the Ryūhonji and Danzan

Shrine versions reveals pictures composed using different techniques and styles. The Ryūhonji

and Danzan Shrine compositions expose vastly different approaches to the innovative

transcription and illustration of the Lotus Sūtra.

Unlike the crowded, even jumbled, composition of Danzan Shrine, the Ryūhonji

mandalas offer a tidy picture surface of distinctly separate vignettes. Large, distilled drawings

capture the narratives of the sūtra and enhance this visual clarity. Because of the larger size of

the vignettes and the lower level of detail offered, the mandalas have a strong legibility. The

195
Ibid., 91.
196
Ibid.
197
Ibid.

106
higher concentration of gold in the paint than that used in the Danzan Shrine set accounts in large

part for the painting‘s preservation. 198 The narrative pictures are more formalized, suggesting

that the tidy individual compositions represent the later techniques used in illustrated sūtra

frontispieces and setsuwa pictures of the thirteenth century. Because of the emphasis on clarity

of composition and distinct visual forms, Miya argues that the content of the narrative vignettes

falls short of those of the Danzan Shrine set.199 The cartouches are also far more abbreviated than

those of the Danzan Shrine version. The simplified and clear composition of bright and engaging

colors, resulting from the brilliant contrast of the luminous gold and dark indigo paper, are

general characteristics of thirteenth-century illustrated sūtra and setsuwa visual culture. This

compositional focus on visual clarity necessitates that scriptural content be omitted, so less

scriptural content is catalogued in the Ryūhonji mandalas. However, because of the tidy layout

and distilled vignettes, the message of the sūtra is perhaps more immediately comprehensible.

Issues of interpretability aside, the Ryūhonji mandalas do embody later compositional technique

and style than the Danzan Shrine paintings, suggesting a later production date of the thirteenth

century.

Style

The intelligibility of the scenes and the overall preservation of the mandalas allow for an easier

analysis of the painting style. While the faces of the Buddhas and devilish creatures exhibit some

strictness, the brushwork is yet soft and fluid as a whole, indicating a date not too late in the

thirteenth century. Indeed, the confident brush renders grass, trees, water, and mountains in the

style of twelfth-century blue and gold sūtra frontispieces. While the painting style of the

198
Ibid., 119.
199
Ibid., 122.

107
narrative vignettes recalls that of twelfth-century illustrated sūtras, the brushwork of the sūtra

characters used to construct the textual reliquary are done with a firmer hand, advocating for a

production date after the start of the Kamakura period. The firmness of the brushstrokes of the

stūpa stands in contrast to the softer, more fluid style seen in the cartouches located on the

mounting at the top of each mandala, which offer the title of the sūtra and volume number. 200

Miya thus suggests that these crowning titles were physically cut from different Lotus Sūtra

volumes produced in the mid-Heian period.201

Whereas the Chūsonji set offers more superfluous landscapes and the Danzan Shrine set

tucks away most of its scenes in mountainous settings, filling in nearly every area of blue paper,

most of the Ryūhonji narratives take place on thin, gold swaths of land. Where more substantial

landscape is used, it is dictated by narrative content. Figures are often allowed to stand alone,

free of heavily contextualized settings. Jettisoning scale, the figures consistently dwarf the

mountains, ravines, and trees. Clearly, the narrative aspect of the vignettes is the focus of the

Ryūhonji mandalas, and to that end the majority of them are clearly marked with succinct

explanatory cartouches. The bodies of the deities and people are drawn in gold with a touch of

red for the lips; this elaborate style is also found in the Danzan Shrine version. The

commonalities in painting style lead Miya to argue persuasively that the Ryūhonji mandalas

inherit the style of the Danzan Shrine paintings, placing the Ryūhonji jeweled-stūpa mandalas in

their lineage. 202 He does, however, clarify that the Ryūhonji version was not produced by

transcription and copying of the Danzan Shrine mandalas—the point is more that Ryūhonji

reflects the Danzan Shrine setsuwa picture tradition.203

200
This is true for all fascicles apart from scroll three.
201
Miya, Kinji hōtō mandara, 90 and 117 n3.
202
Ibid., 122.
203
Ibid.

108
 Both the technique and style of the Ryūhonji paintings suggest a date later than the

Danzan Shrine version‘s production, estimated ca. 1187. Visual evidence in fact encourages a

date not later than the mid-thirteenth century. And while the mandalas cannot be situated in the

context of a commission which would further inform the production date as in the case of the

other two sets, by considering the earliest known restoration of the mandalas, it is possible to

gauge their time of creation. Based on the general trends in the maintenance of images in the

medieval period, most initial restorations were made one hundred years after the object‘s initial

production.204 Applying this hypothesis to the Ryūhonji jeweled-stūpa mandalas and calculating

back from the earliest recorded mending of 1362 would date the paintings to the mid-thirteenth

century, a time substantiated by the visual qualities of the mandalas.

Formal Analysis

Process of construction

Eight fascicles compose the Ryūhonji jeweled-stūpa mandala set, offering the entirety of the

Lotus Sūtra‘s twenty-eight chapters grouped traditionally in eight volumes, one volume per stūpa.

The Ryūhonji paintings are beautifully preserved. Nearly every scene remains vibrant and

complete, the still-glittering gold paint contrasting brilliantly against the deep blue background.

While the Chūsonji set is minimalistic and the Danzan Shrine version cramped, the Ryūhonji

mandalas are perhaps the most balanced. Despite the more simplistic and condensed narratives,

Ryūhonji‘s mandalas still conform to the characteristics of transformation tableaux depicting the

tales of the twenty-eight chapters of the scripture.

The production of the Ryūhonji mandalas presents the same challenges and creative

opportunities as the other two sets. In accordance with the process of jeweled-stūpa mandala
204
Ibid., 116.

109
construction, a critical planning stage for such complicated compositions was conducted with the

Ryūhonji set. Given the precision of the stūpa construction, shallow, engraved lines made by a

metal stylus dictated the location of each line of scriptural text before even a character was

brushed. And much like the other two sets, each mandala is crafted using seven sheets of paper

dyed a rich blue. The central stūpa is composed of two narrow sheets joining together at the

fourth story of the reliquary. Immediately out from this same spot into the space occupied by the

vignettes on the right and left of the stūpa, two sheets join to make the bordering narrative bands.

The seventh sheet attaches to the mandala directly underneath the foundation of the architecture

and presents the lowest narrative scenes. The joints at the convergences of the seven papers are

virtually indistinguishable as they masquerade as lines of architecture or details of landscape—

only the two, long seams running the vertical distance of the mandala are visible. Unlike the

Danzan Shrine mandalas, the dyed blue color is largely consistent in hue throughout the

Ryūhonji paintings. However, given the procedures I have hypothesized concerning the order of

production with the other two mandalas, it is still likely that the Ryūhonji mandalas witnessed

the same assembly line style construction. The incursion of the narrative vignettes into the realm

of the textual reliquary obliges the transcription of the stūpa before the narratives were

completed. Such a laborious process surely required the skills of multiple trained painters of

Buddhist subjects and copyists of Buddhist texts. The Ryūhonji paintings also carry another

signature of the jeweled-stūpa mandala category: that of the arabesque frame bordering all sides

of the mandala composition. The primary stems of the elaborate and interweaving flora pattern

are rendered in gold, while the delicate leaves shine in silver.

Stūpa

110
Growing out of thinly washed swaths of golden land on the edge of a silver sea, 205 the stūpa of

the Ryūhonji mandalas provides a central balance for the composition of narrative vignettes. The

description of the reliquary generally matches the stūpas used in the jeweled-stūpa mandala

category of painting, although, as already described in the above sections, the stūpas of each set

are characterized by individual characteristics. Ryūhonji‘s stūpa depicts steps leading up to the

platform enclosed by a low banister. The description of those steps is distinct to each set in terms

of character alignment and spacing. Both Ryūhonji and Chūsonji ground their stūpas with solid

foundations and stairs that flare out to the right and left at the bottom of the dais, thus presenting

the stūpas as accessible monuments. Chūsonji‘s steps are composed of evenly dispersed,

vertically oriented characters whose wide horizontal spacing prevents their appearance as a

viable pathway. Quite the opposite, Ryūhonji‘s stairs read as plausible steps, achieving visual

closure through tightly packed, horizontally oriented characters forming parallel and consistent

lines. Unlike the other two sets, Ryūhonji‘s golden stūpa does not rest on a lotus pedestal. The

stūpa of the Ryūhonji set also reads as more transitory, for while the key architectural features

are represented and the stūpa itself is gently grounded in golden land, at several points on the

stūpa, the blue sea of silver waves is visible through the reliquary. For instance, the railed

platform spread before the opened doors of the stūpa is not described in color like the Chūsonji

mandalas nor composed of sūtra characters as is the case in the Danzan Shrine version. The same

is true of the exterior wall panels flanking the door of the reliquary—again bright blue ocean

glimmers through the supposedly solid architecture, causing the viewer to acknowledge the

idiosyncrasies of a stūpa constructed by textual characters. Inside the stūpa, two Buddhas

identical in appearance sit side-by-side. This scene would be instantly recognizable as the

common iconography of Śākyamuni and Prabhūtaratna from the ―Apparition of the Jeweled-
205
The waves of the sea are depicted with silver paint in all but the first and second fascicles.

111
Stūpa‖ chapter of the Lotus Sūtra in which Prabhūtaratna appears in a glorious stūpa while

Śākyamuni preaches the sūtra, in another meta-reference in the scripture:

At that time, there appeared before the Buddha a seven-jeweled stūpa, five hundred
yojanas [a distance of either seven or nine miles; 由旬 Jpn. yujun, Ch. youxun] in height
and two hundred and fifty yojanas in breadth, welling up out of the earth and resting in
midair, set about with sundry precious objects. It had five thousand banisters, a thousand
myriads of grottolike rooms, and numberless banners to adorn it. Jeweled rosaries trailed
from it, and ten thousand millions of jeweled bells were suspended from its top.
Tamālapatracandana [name of a fragrance and a Buddha; 多摩羅跋栴檀之香 Jpn.
Tamara sendan no kō, Ch. Duomoluo zhantan zhi xiang] scent issued from all four of its
surfaces and filled the world; its banners were made of the seven jewels, to wit, gold,
silver, vaiḍūrya [lapis lazuli; 琉璃 Jpn. ruri, Ch. liuli], giant clamshell, coral, pearl, and
carnelian; and its height extended to the palaces of the four god kings. The thirty-three
gods rained down on it divine māndārava flowers [a red heavenly flower; 曼陀羅華 Jpn.
mandara ke, Ch. mantuoluo hua], with which they made offerings to the jeweled stūpa.
the other gods, dragons, yakṣas [devilish lower-ranking deities; 夜叉 Jpn. yasha, Ch.
yecha], gandharvas [centaur-like lower-ranking deity; 乾 闥 婆 Jpn. kendatsuba, Ch.
gantapo], asuras, garuḍas [winged lower-ranking deity; 迦 樓 羅 Jpn. karura, Ch.
jialouluo], kiṃnaras [heavenly musician part human, part animal 緊那羅 Jpn. kinnara,
Ch. jinnaluo], mahoragas [great snake spirit; 摩 睺 羅 伽 Jpn. magoraga, Ch.
mohuoluoqie], humans, and nonhumans, numbering a thousand myriads of millions,
made offerings to the jeweled stūpa of all manner of flower perfumes, necklaces, banners,
and skillfully played music, reverently worshiping it, holding it in solemn esteem, and
singing its praises. At that time, from the midst of the jeweled stūpa issued forth the
sound of a mighty voice, praising and saying, ―How excellent! How excellent, O
Śākyamuni, O World-Honored One, that with great undifferentiating wisdom you can
teach the bodhisattva-dharma, that you can preach to the great multitude the Scripture of
the Blossom of the Fine Dharma, which buddhas keep protectively in mind! Verily,
verily, O Śākyamuni, O World-Honored One! Whatever you preach is all true reality.‖ 206

The lavish detail describing the greatness of the jeweled-stūpa illustrates the prominence of the

reliquary and its role as an icon to be sumptuously worshiped.

The stūpa is nine-stories and lacks the false roof seen in the Chūsonji and Danzan Shrine

versions. The main bracketing system is appropriately depicted with supporting bases, rafters and

arms attaching beams to pillars, and the window of each floor showing light green latticing,

framed in red. Once again, the principle architectural components of the finial are clearly drawn.

206
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 167. T. no. 262, 9: 32b17-32c2.

112
Starting from the base of the finial, the seven common features are described: (1) the square box

supporting the weight of the finial, (2) the inverted bowl which supports the flower base, (3) the

simple flower base which hosts the above structure, (4) the nine metal rings decoratively

attached to the central pillar, (5) the metal decoration affixed to the exposed pillar, (6) the oval

element between the jewel and the metal decoration below, and (7) the crowning jewel.

The arduous process of building each stūpa from diminutive sūtra characters begins with

the uppermost character of the precious jewel adorning the finial. From this highest character, the

text travels down each story of the reliquary and ends at the bottom of the front stairs on the

right.207 Much like the Danzan Shrine mandalas, an attempt is made to end each stūpa with the

last characters of the scripture followed by the explanatory attachment indicating the title of the

sūtra and the volume number. Scroll one and two end perfectly as planned. 208 However, the

transcription becomes more complicated after this. Volumes three, five, seven, and eight lack the

length required to construct the large reliquary and so verses are attached to the conclusion of the

last chapter which is then appended by the sūtra title and volume number. 209 Volume four adds

an abridged title of the sūtra to the end of the eleventh chapter: Myōhokekyō 妙法華経.210

Battling the opposite transcription challenge, volume six is too long to fit completely and so the

remainder is omitted and concluded with the same formula of sūtra title and volume number. 211

Narrative Vignettes

This divine realm of indigo and gold where word constructs architecture and deities and humans

intermingle casts a striking vision of the potentialities of the Lotus Sūtra. Perhaps it is because

207
Miya, Kinji hōtō mandara, 91.
208
Ibid.
209
Ibid. For a transcription of the verse used, see Ibid., 117 n6.
210
Ibid., 91.
211
Ibid.

113
the Ryūhonji jeweled-stūpa mandalas are made of high-quality gold and remain so well-

preserved that they appear as the strongest contrast of deep blue and brilliant gold. Indeed, the

mandalas stand as a golden realm, affected so partly because gold is the primary color whereas

the Chūsonji set displays a rich palette and the severely faded Danzan Shrine set offers a more

muted gold balanced by a profusion of silver. This visual formula derived from the blue and gold

illustrated sūtra also differs from the artistic techniques witnessed in the handscrolls. While the

animals are described in silver, all the Buddhas, principles bodhisattvas, people, and even

demons are rendered in gold. The eyes, nose, hands, and legs are detailed with black ink and the

mouths of the Buddhas, bodhisattvas, and people are dotted with red. This description closely

resembles the techniques adopted in the Danzan Shrine mandalas. While the silver description of

the buildings‘ walls and the ephemeral mist tempers the abundance of gold, the world of the

jeweled-stūpa mandalas remains a golden one.

In addition to their vibrancy, because the Ryūhonji mandalas are the widest of the three

sets and because the scenes themselves are larger and selectively chosen rather than

comprehensively presented, the paintings are much easier to read. As noted before, the Chūsonji

scenes are not so much narratives as images of symbolically charged deities and worshipers.

However, with its endless parables and episodes, the Lotus Sūtra readily lends itself to

pictorialization. Thus, the Ryūhonji and Danzan Shrine sets are replete with narrative scenes.

Approximately one hundred and twenty illustrations populate the eight mandalas of Ryūhonji. 212

The composition is deftly proportioned and avoids overwhelming the viewer, a tendency likely

to occur with the Danzan Shrine version. While certainly not a steadfast rule, the narrative

vignettes tend toward an order placing the chapters occurring earlier in the volume in the lower

half or right of the mandala; from this starting point, the narratives travel upwards to the top of
212
Ibid., 114.

114
the right band before switching over the lower left and working their way up to the top of the left

band. 213 This configuration technique follows the conventional Japanese narrative arrangement

style, such as that seen in the Shōtoku Taishi biographical painting. 214 The distilled composition

of the Ryūhonji mandalas offering roughly half of the vignettes of Danzan Shrine—a set more

directly connected with the continental models—represents a transition toward a more Japanese

style of composition, and particularly, a thirteenth-century composition. The privileging of tidy,

bright, readable scenes as well as the fewer and sparser cartouches signify a shift toward the

styles popular in thirteenth-century Japan. The Ryūhonji mandalas intersperse cartouches around

the stūpa offering passages from the scripture, typically with between seven to twenty-one

cartouches attached to each scroll with a total of one-hundred and six—far fewer than those

offered in the Danzan Shrine set.215 These quotes are often much shorter and occasionally the

ambiguity and brevity of the passage prevents an easy identification of the scene on its own. But,

owing to the clarity of the graphic episodes, the intended scriptural content remains accessible.

Furthering the theme of thirteenth-century pictorial inclinations observable in the

Ryūhonji jeweled-stūpa mandalas, the limited number of vignettes requires a highly selective

process in determining which narratives to illustrate, a process that privileges popularity over

comprehensiveness. A survey of the scenes illustrated in each mandala 216 reveals that stories

with themes amenable to visual description as well as stories embedded in time requiring

continuous narration were the ones most frequently chosen. By examining and collating the most

frequently depicted scriptural episodes on blue and gold illuminated sūtras copied between the

eleventh through thirteenth centuries, Miya proposes a ranking of the most popular narrative

213
Ibid., 120.
214
Ibid., 135.
215
Interestingly, Miya privileges the density of the Danzan Shrine mandalas over the tidily composed Ryūhonji
paintings, explaining that he feels the Ryūhonji version lacks the intensity of the Danzan Shrine set. See Ibid., 121.
216
Ibid., 112-14 lists the scenes depicted in each painting.

115
images.217 These popular images correspond directly to the choices made in the narrative

selection process of the Ryūhonji mandalas and moreover, occur in particular spaces around the

textual stūpa. By charting the arrangement of the scenes, Miya also discovers that the most

popular and beloved episodes from the Lotus Sūtra were concentrated in the top and bottom of

each mandala, 218 with the exception of the bottom of scroll one 219 which depicts the standard

scene of preaching on Vulture Peak. The bottom of scroll two 220 offers a rather large scene of the

popular burning house parable. 221 The upper right of the scroll continues one of the most

important lessons from the Lotus Sūtra: the three carts stationed outside the burning house are

revealed to be metaphors for expedient means (方便 Jpn. hōben, Ch. fangbian; Skt. upāya) to

nirvāṇa based on the practitioner‘s abilities. The three vehicles (三乗 Jpn. sanjō, Ch. sansheng;

Skt. triyāna) are the auditors (聲聞 Jpn. shōmon, Ch. shengwen; Skt. śrāvaka), individually

enlightened (緣覚 Jpn. engaku, Ch. yuanjue; Skt. pratyekabuddha), and bodhisattva. The Lotus

Sūtra preaches that this form of understanding reflects a preliminary time, and that in the Lotus

Sūtra all three means are subsumed, offering the sūtra as the one Buddha vehicle (一仏乗 Jpn.

ichibutsu jō, Ch. yifo sheng; Skt. ekayāna) or perfect path to buddhahood. Scroll three 222 also

continues this trend by placing a scene from the ―Medicinal Herbs‖223 chapter in the lower right-

hand corner of a substantial palatial garden with royal attendants. This chapter again explains the

217
Ibid., 115.
218
I choose to highlight a sampling of the scenes to illustrate this point. For the full list, see Ibid., 114-15.
219
This scroll measures 111.4 cm x 58.5 cm and includes chapter one, ―Introduction,‖ and chapter two, ―Expedient
Devices.‖
220
Scroll two measures 111 cm x 58.7 cm and includes chapter three, ―Parable,‖ and chapter four, ―Belief and
Understanding.‖
221
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 55-60. T. no. 262, 9: 12b21-13c18.
222
This scroll measures 111.2 cm x 58.5 cm and includes ―Medicinal Herbs,‖ ―Bestowal of Prophecy,‖ and ―Parable
of the Conjured City.‖
223
Jpn. Yakusō yu, Ch. Yaocao yu; 藥草喩品.

116
concept of expedient means using the metaphor of medicinal herbs. 224 The fourth scroll225 offers

in the upper left of the mandala one of the most popular scenes from the Lotus Sūtra, that of the

apparition of the Tahō stūpa.226 Substantial space is devoted to the depiction of this scene. Above

a crowd of worshiping people, a jeweled stūpa hovers and reveals a seated Tahō. Deities on

clouds surround the apparition. The upper left of scroll six 227 presents the dynamic image of

warring demons in a palace by the edge of the great sea.228 This scene comes from ―The Merits

of the Dharma Preacher‖ chapter which proclaims the many rewards gracing those who uphold

the scripture, including the ability to hear the sounds of warring demons talking in their remote

abode.229 The scene depicting the fall from a diamond mountain of one chased by an evil man is

represented in the lower right corner of the eighth mandala. 230 This scene is part of a larger series

of narratives depicting Kannon‘s intercessory salvations of those in trouble and of the earnest in

prayer which occupies the majority of the eighth scroll, much like that of Danzan Shrine.

Lone Jeweled-Stūpa Mandalas

At least two jeweled-stūpa mandalas, separated from their now lost sets, have surfaced in recent

years. One of these mandalas has a proposed early twelfth-century date and is housed in a private

collection. Unfortunately, this particular mandala has not fared well and many of the narrative

224
For example, see Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 95-6. T. no. 262, 9: 19a27-19b21.
225
This scroll measures 111.5 cm x 58.7 cm. This scroll contains the fourth volume of the Lotus Sūtra, including the
chapters: ―Receipt of Prophecy by Five Hundred Disciples‖ as chapter eight, ―Prophecies Conferred on Learners and
Adepts‖ as chapter nine, ―Preachers of the Dharma‖ as chapter ten, and ―Apparition of the Jeweled-Stūpa‖ as
chapter eleven.
226
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 167. T. no. 262, 9: 33b26-28.
227
Scroll six measures 111.2 cm x 58.7 cm. This scroll includes ―The Life Span of the Thus Come One‖ as chapter
sixteen, ―The Merits of Appropriate Joy‖ as chapter eighteen, and ―The Merits of the Dharma Preacher‖ as chapter
nineteen. Chapter seventeen, ―Discrimination of Merits,‖ is not depicted.
228
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 244. T. no. 262, 9: 48a21-22.
229
Jpn. Hosshi kudoku bon, Ch. Fashi gongde pin; 法師功徳品.
230
This mandala measures 110 cm x 58.8 cm. and includes chapter twenty-five, ―The Universal Gateway of the
Bodhisattva He Who Observes the Sounds of the World,‖ chapter twenty-six, ―Dhāraṇi,‖ and chapter twenty-seven,
―The Former Affairs of the King of Fine Adornment.‖ Chapter twenty-eight, ―The Encouragements of the
Bodhisattva Universally Worthy,‖ is not represented.

117
vignettes are barely legible. The other example is stored in Myōhōji with a suggested date of the

late twelfth century or early thirteenth century. The Myōhōji mandala survives in great condition

with gold still radiant and the vignettes clear and readable. Despite the disconnectedness of these

paintings from their original sets, these two examples are very valuable because they establish a

wider production of the jeweled-stūpa mandalas at this time than previously thought.

Jeweled-Stūpa Mandala in a Private Collection

How this particular jeweled-stūpa mandala came to be in the possession of a private collector is

not currently publicized. The original commission is also a mystery, leaving very little to

understand the context of this mandala‘s production. However, a different jeweled-stūpa

mandala of striking similarity to this privately owned example surfaced in the late 1990s at

Jōshinji 浄真寺 in Shiga prefecture. The compositional structure, stūpa construction, and

rendition of the narrative vignettes are so very alike that it is highly probable that they were

originally from the same set and later separated.231 Unfortunately, to my knowledge, nothing of

substance has been published yet on the Jōshinji mandala, including high quality photographs.

The privately housed mandala transcribes the sixth volume of the Lotus Sūtra.232

Unlike the faded composition surrounding the reliquary, the gold of the stūpa gleams.

This is because only the central icon was restored. This is also characteristic of the Jōshinji

mandala. The stūpa of this transcription is far simpler than those of the others mandalas. Like the

Chūsonji and Danzan Shrine versions, this reliquary is nine-levels with a false roof. This

mandala also terminates in a squared foundation without handrails and steps, like the Danzan

231
For a small image of the Jōshinji mandala, see Kyoto National Museum, Ōchō no butsuga to girei zen o tsukushi
bi o tsukusu 王朝の佛画と儀礼: 善をつくし美をつくす (Kyoto: Kyoto National Museum, 1998), 343. The
Jōshinji example presents the third volume of the Lotus Sūtra.
232
This scroll is slightly smaller than the other mandalas, measuring 96.2 cm x 46.5 cm. For an image of this
mandala, see Izumi, ―Hokekyō hōtō mandara,‖ 29.

118
Shrine set. However, this mandala lacks the decorative flourishes seen in the other versions. The

transcription also reads as more cramped and the characters are indistinct from one another.

Typical of jeweled-stūpa mandalas, the scripture begins at the top of the finial and concludes in

the right corner of the foundation with the title of the sūtra and volume number. Housed inside

the reliquary are two seated Buddhas, now faint in description and barely visible. While gold is

the most obvious ink used, silver accents can still be detected in details of the stūpa, like the

paneling and windows, and in the landscape and buildings of the narrative vignettes.

Unfortunately, the extensive fading of the vignettes makes discernment of most scenes

nearly impossible. 233 Fewer scenes are offered overall. The arrangement of the narratives

proceeds much like the Chūsonji vignettes: beginning in the top right and continuing to the

bottom of the column then crossing over to the top left column and continuing down. 234 In his

analysis of the mandala, Izumi Takeo concludes that based on the style of the Buddhas,

bodhisattvas, and their clothing which reflects a softness and roundness seen in the late eleventh

century and early twelfth century, the mandala was produced no later than the early twelfth

century, making this painting the oldest example of the Japanese jeweled-stūpa mandalas.235

Myōhōji’s Jeweled-Stūpa Mandala

Sadly, the same lacuna plaguing each example of this rare form of painting haunts the only

remaining fascicle from a lost set now housed at Myōhōji in Sakai. 236 Before the 1985 exhibition

in Sakai City Museum 堺市博物館 entitled ―Sakai no butsuzō butsuga 堺の仏像仏画,‖ few

233
For a description of what can be seen of the narratives, see Ibid., 29-38.
234
Ibid., 36.
235
Ibid., 35, 37. Interestingly, Izumi disagrees with Miya‘s hypothesis that the high density of the Danzan Shrine
narratives embodies an earlier style while the pared down quality of the Ryūhonji vignettes reflects a later style.
Instead, Izumi suggests reversing the order, with Danzan Shrine manifesting the later style and Ryūhonji recapturing
an earlier, sparser style, like that of the privately owned mandala. For more on this argument, see Ibid., 37-38.
236
For an image of this painting, see Kyoto National Museum, Koshakyō, 171 plate 94.

119
people knew of the existence of this lone jeweled-stūpa mandala. Nothing is known of its

commission, original set context, or historical trajectory, for instance, how it came to be stored at

Myōhōji. Curiously, the swirling legends of the apotheosized Shōtoku Taishi, continuing to grow

in prominence long after the death of the real prince, touched even this mysterious mandala. An

inscription on the back of the scroll claims that the brush of the great prince composed the

mandalas.237 No explanation of such a boldly spurious claim is offered either on the mandala

itself or in any known records. The inscription also describes the dedication of the ‗ten-

storied‘238 mandala on an auspicious day in the ninth month of 1641 by the monks Nichiyō 日遙

of Ryūhonji and Nichitō 日東.239 This is a curious connection between two of the few temples to

acquire a jeweled-stūpa mandala. However, according to textual records discussed in the above

section on Ryūhonji‘s set history, the mandalas appear to have been in the possession of Hōryūji

until at least 1652 before they were transferred to Ryūhonji in 1681. Therefore, the idea that

perhaps Nichiyō was requested to the dedication service because of the temples‘ shared rare

objects would not apply. Instead, the request likely stems from the temples‘ shared school

affiliation as Ryūhonji and Myōhōji were both Nichiren temples. Perhaps though, the exposure

of Nichiyō to the jeweled-stūpa mandala at Myōhōji brought this unusual style of Lotus Sūtra

ornamental transcription to the attention of Ryūhonji at an advantageous time as Hōryūji‘s

mandalas would be available only forty years later, thus making it possible that this category of

painting would still be in the collective temple memory.

The Myōhōji mandala is the seventh from a series of either eight or ten paintings,

237
Miya, ―Myōhōjizō myōhōrengekyō kinji hōtō mandara ni tsuite,‖ 96 n1.
238
The central architecture of the Myōhōji mandala is a nine-storied structure with a false roof, much like the
Chūsonji and Danzan Shrine versions and the mandala held in a private collection.
239
See inscription in Miya, ―Myōhōjizō myōhōrengekyō kinji hōtō mandara ni tsuite,‖ 96 n1.

120
depending on whether or not the opening and closing sūtras were included. 240 The construction

of the mandala resembles the standard techniques found in the other mandalas. As best as I can

discern using reproductions of the Myōhōji painting, the mandala is constructed of seven

separate pieces of indigo dyed paper. Two sheets containing the narrative vignettes run the

length of both sides of the stūpa, joining at the seventh roof from the top. The central textual

stūpa is composed of two sheets of paper also attached at the joint of the seventh roof. While

difficult to see clearly, it appears that the narratives below the foundation of the stūpa are also

painted on a separate sheet.

Used for writing the majority of the stūpa, the bodies of most of the figures, and the

landscape and building, gold dominates the composition. However, silver accents are also

profuse throughout the mandala. Many of the human faces and a few of the deities‘ bodies are

rendered in silver as well as washes along the landscapes, details of the buildings, and parts of

the stūpa such as the paneling, the forward-facing foundation, and the steps which alternate gold

and silver. As noted in the sections analyzing the Danzan Shrine and Ryūhonji sets, the

technique of using gold and silver for the bodies of the narrative figures is rarely found in

Japanese blue and gold illuminated sūtras, suggesting that this is another hallmark of the style of

the jeweled-stūpa mandalas.

Unlike the Danzan Shrine set and the majority of the Ryūhonji mandalas, the golden

stūpa of Myōhōji does not emerge out of an ocean of waves, and in this way, the Myōhōji

version resembles the Chūsonji set. Much like the Ryūhonji and privately owned mandalas,

seated inside the stūpa are two Buddhas painted in gold and heavily damaged. A handrail

surrounds the foundation and leads down the steps, again manifesting the same architectural style

as the Chūsonji and Ryūhonji mandalas. Typical of the Japanese jeweled-stūpa mandalas, the
240
This scroll measures 103.7 cm x 54.5 cm.

121
sūtra transcription begins at the top of the finial and ends with the last step of the stūpa,

terminating at the curve of the right handrail.

The narrative vignettes of the Myōhōji mandala depict many of the conventional

illustrations from the seventh volume of the Lotus Sūtra. The favorite vignettes are represented,

and like Ryūhonji, the arrangement of the scenes is less orderly with the narratives being more

randomly distributed. The explanatory cartouches also resemble those of the Ryūhonji version,

except that the Myōhōji painting offers complete titles and excerpts, whereas the Ryūhonji

mandalas occasionally omit parts.241 Given the similarities with the Ryūhonji set in terms of the

architectural style of the stūpa, the scene selection and arrangement of the narratives, and the

mirroring of the cartouches, compounded by the late twelfth-century style of the paintings

reflected in the brushwork, the gentle facial expressions of the figures, and the treatment of the

landscape, Miya proposes that painters and copyists of the Ryūhonji set used the Myōhōji

mandala as a model. 242 He also suggests that the lack of great similarities with the Danzan Shrine

and Chūsonji versions indicates no direct relationship.243 Based on these observations, a date of

the late twelfth or early thirteenth century has been posited.244

Conclusion

Examining the circumstances of the commission and formal characteristics of the jeweled-stūpa

mandalas of the Chūsonji, Danzan Shrine, and Ryūhonji sets, along with the two lone mandalas,

reveals the singularity of this category of paintings and, at the same time, their indebtedness to

continental models and Japanese blue and gold illuminated sūtras. This chapter explored the

241
Miya, ―Myōhōjizō myōhōrengekyō kinji hōtō mandara ni tsuite,‖ 89-90.
242
Ibid., 94-96.
243
Ibid., 94-95.
244
Ibid.,96.

122
challenging construction process for the mandalas and approximate production dates were

offered based on examinations of the techniques, styles, and commission context of the painting

sets. Chūsonji‘s paintings offer an opportunity to examine the private side of jeweled-stūpa

mandala commissions, revealing a matrix of confession, anxiety, ambition, and ingenuity. The

importance of the Golden Light Sūtra to the Ōshū Fujiwara and their concerns, on the one hand,

for strong authoritarian rule and, on the other, for personal salvation is manifested in the

innovative transcription and illustration in the mandalas. The mandalas were shown to be the

products of a final commission by Fujiwara Hidehira in 1170. The Danzan Shrine set offers more

in-depth visual descriptions of the Lotus Sūtra’s parables perhaps due to the mandala‘s potential

commission in 1187 to commemorate the priest Zōga and his celebrated love for the scripture.

The cramped composition and narrative vignette structure speaks to the mandalas‘ continental

connections. The Ryūhonji set, a study in balance and artistic control, reflects the distillation of

sūtra illustrations and cartouches which, combined with the tidy and bright composition,

embodies the visual style and technique of thirteenth-century scriptural imagery. This

approximate date is further narrowed down by calculating one century back from the earliest

recorded restoration in 1362, placing the production date of the mandalas sometime around the

mid-thirteenth century. A thorough visual and contextual analysis is important in establishing a

sound basis on which an interpretive theoretical framework might stand. To this end, the

following three chapters move from practical issues and concerns focused on the specifics of

each mandala and each set to the conceptual interpretation of the jeweled-stūpa mandalas as a

whole.

123
 Chapter Four

Dharma Reliquary and the Power of Word

Introduction

Text is more than just inscribed letters. Indeed, whether it is word etched on memory, a

vocalized mantra (literally ‗true word,‘ esoteric chant; 真言 Jpn. shingon, Ch. zhenyan), or even

the entire universe, text need not be limited to written script. The profundity and numerous

manifestations of text in early medieval Japan represent the integrality of word to discourse, to

political and cosmic authority, and to uncovering the reality masked by illusion. The power of

word saves lives and redeems souls, spurs creative and elaborate statements about the nature and

potentialities of text in visual culture, and—most important to this study—grounds the jeweled-

stūpa mandala‘s meaning.

I begin this chapter by establishing the power of text and dharma relics (法舎利 Jpn.

hōshari, Ch. fa sheli; Skt. dharma śarīra)1 as conceived in early medieval Japan and by

examining the ubiquitous practice of copying scriptures as a means of harnessing this textual

benefit. I proffer the mandalas as an elaborate example of such practice and maintain that it is the

inherent power of sacred word that acts as a catalyst for the mandala‘s creation. I then pursue the

combinatory practices merging sūtra and stūpa as a precedent for the elaboration witnessed in the

jeweled-stūpa mandalas.

1
Dharma relics are further revealed to be the dharmakāya (dharma body; 法身; Jpn. hōshin; Ch. fashen) of the
Buddha, a subject examined in the following chapter on the bodies of the Buddha.

124
Power of Sacred Word

As has already been introduced in the first and second chapters, the lives of early medieval

sacred texts were many and diverse. Venerated for their inherent salvific power, coveted for their

social and economic cachet, possessed for their authentication of political and religious authority,

sūtra texts enjoyed a central position in early medieval Japanese textualized society. But how can

we conceive of this all-encompassing power? And more to the point, what is it exactly? Sasaki

Kōkan 佐々木宏幹 has analyzed the power inherent in sūtras and avenues of access in his study

considering the relationship between the monastic community and the emperor in regards to

spiritual power.2 He defines this scriptural power as juryoku 呪力 (power of sūtras and dhāraṇīs),

which is accessed through a variety of ceremonies involving chanting and copying the sūtras. 3

Those who wield the power of sūtra are termed jushi 呪師 (‗shaman‘) and have considerable

influence with the imperial family and aristocrats.4 The activity of manipulating scriptural power

for particular benefits is jujutsu 呪術 (‗incantion‘ or mantra).5 Jujutsu is fundamentally the

establishment of two previously unrelated phenomena, one which operates through its power on

the other to achieve a certain result. 6 The transference of power from the sūtra to the intended

object through the operator (jushi) establishes a store of power within the operator of the sūtra,

which originates from the contact with the scriptural power and continues to exist in the wielder

of sūtra.7 In the cases of court sponsored rituals and ceremonies, the power is stored within

ordained monks. However, in instances where an individual performs a ceremony, such as sūtra
2
Sasaki Kōkan 佐々木宏幹, ―Sō no jushika to ō no saishika: bukkyō to ōsei to no musubitsuki ni kansuru ichi
shiron 僧の呪師化と王の祭司化: 仏教と王制との結びつきに関する一試論,‖ in Kokka to tennō tennōsei
ideorogi to shite no bukkyō 国家と天皇 天皇制イデオロギ-としての仏教, ed. Kuroda Toshio 黒田俊雄 (Tokyo:
Shujūsha, 1987), 49-91.
3
Ibid., 53.
4
Ibid.
5
Ibid.
6
Nakamura, Iwanami bukkyō jiten, 407-08.
7
Sasaki, ―Sō no jushika to ō no saishika,‖ 63.

125
transcriptions for the benefit of deceased loved ones, the power of the scripture is transferred to

the intended object, yet also resides in the practitioner. The manipulation of scriptural power

secures ambitions as grand as eternal salvation, political success, protection from harm, and

health and longevity; and as mundane as temporal wishes for good harvests and the all-important

control over rain. Sūtras are greater than their materiality and orality; they are imbued with a

dynamic, sacred power that serves as efficacious talismans. If harnessed, great miracles and

rewards await.

Understanding this power in explicit terms with historical and doctrinal certitude is

unlikely. But in an effort to qualify the use of such a nebulous and definitionally elusive term as

power and offer an explanation of this potency of sacred word and its basis as a catalyst for the

mandalas, I approach the subject from a phenomenological perspective. Therefore, what follows

is an examination of sūtras‘ self-referential boasts of their inherent powers, claims in setsuwa

and aristocratic diaries about scripture‘s miraculous abilities, and monastic commentaries on

sacred word‘s potentialities. I then investigate the origin of this scriptural power as dharma relic

and offer examples in religious practice—such as copying and burying sūtras—with doctrinal

and visual support as further evidence of this phenomenon. The jeweled-stūpa mandalas are

revealed to be a prime illustration of the equivalence of sūtra, dharma relics, and Buddha.

Testaments to Their Own Abilities

More than just vessels and vehicles for access to salvific power and enlightenment, sūtra texts

embody soteriological potential as material manifestations of buddhadhātu (Buddha-nature).8

Because a vast number of sūtras testify self-referentially to the limitless capabilities of sacred

8
The nonduality of sūtra text as dharma relics and dharmakāya and buddhadhātu is analyzed in the following
chapter.

126
word, I shall not undertake to catalogue each sūtra‘s proclamations. 9 To this effort, I consider the

Lotus Sūtra, the Golden Light Sūtra, and the Perfection of Wisdom Sūtras, only a few of the

numerous scriptures which declare sūtra‘s manifest power to assist the user of the text in sundry

ways.10 The goal of this section is to demonstrate the doctrinal justification for the active and

salvific power imbued in sacred word by sampling sūtras that advocate directly and forcefully

the power invested in scripture. I also want to make the case that the sūtras employed in the

jeweled-stūpa mandalas are texts that have strong elements of sūtra worship. This assertion is

then further developed in the chapter to suggest that this display of the scriptures is the visual

manifestation of sacred word as dharma relics in a form that signifies the worship of

Buddhas/relics and expresses the multivalence of Buddha body doctrine.

The Lotus Sūtra, not unlike other scriptures, proclaims itself to be the most important text

in the Buddhist canon. In chapter eleven, ―Apparition of the Jeweled Stūpa,‖ when summarizing

the important lessons of the chapter, the Buddha says:

For the sake of the Buddha path, I,

In incalculable lands,
From the beginning until now,
Have broadly preached the scriptures,
But among them

9
For instance, the Flower Garland Sūtra visualizes the universe textually. See Luis O. Gómez, ―The Whole
Universe as a Sūtra,‖ in Buddhism in Practice, ed. Donald S. Lopez (Princeton: Princeton University Press, 1995),
107-112.
10
Conspicuously absent from this discussion of the power of sacred word are mantras (真言; Jpn. shingon; Ch.
zhenyan) and dhāraṇīs (陀羅尼; Jpn. darani; Ch. tuoluoni). While I touch on these concepts later in the chapter and
in the sixth chapter when discussing the orality of scripture, a detailed examination falls outside the borders of this
project given the focus on the jeweled-stūpa mandalas. For thorough analyses of mantras and dhāraṇīs, see Abé, The
Weaving of Mantra; Donald S. Lopez, ― Inscribing the Bodhisattva's Speech: On the ‗Heart Sūtra's‘ Mantra,‖
History of Religions 29 (1990): 351-72; Donald S. Lopez, Elaborations on Emptiness: Uses of the Heart Sūtra
(Princeton: Princeton University Press, 1996); Richard K. Payne, ―Awakening and Language: Indic Theories of
Language in the Background of Japanese Esoteric Buddhism,‖ in Discourse and Ideology in Medieval Japanese
Buddhism, ed. Richard R. Payne and Taigen Dan Leighton (London and New York: Routledge, 2006), 79-96; James
Robson, ―Signs of Power: Talismanic Writing in Chinese Buddhism,‖ History of Religions 48 (2008): 130-69; and
Glenn Wallis, ―The Buddha‘s Remains: Mantra in the Mañjusrimulakalpa,‖ Journal of the International Association
of Buddhist Studies 24 (2001): 89-125.

127
 This scripture is first.11

By proclaiming the supremacy of the sūtra, the scripture positions itself as worthy of worship

and, with the proper attention and devotion, followers of the sūtra are promised access to the

many rewards of the scripture. At several points the sūtra instructs devotees to copy and recite its

text, venerate its rolls with offerings, and disseminate the dharma, resulting in great rewards for

the practitioner, such as the direct protection of the Buddha: ―O Medicine King, be it know that

after the extinction of the Thus Come One, those who can write it, hold it, read and recite it,

make offerings to it, or for others preach it the Thus Come One shall cover with garments.‖ 12

The scripture also promises that those who uphold the sūtra will be accorded the honor and gifts

of a Buddha:

[I]f a good man or good woman shall receive and keep, read and recite, explain, or copy
in writing a single phrase of the Scripture of the Dharma Blossom, or otherwise and in a
variety of ways make offerings to the scriptural roll with flower perfume, necklaces,
powdered incense, perfumed paste, burned incense, silk banners and canopies, garments,
or music or join palms in reverent worship, that person is to be looked up to and exalted
by all the worlds, showered with offerings fit for a Thus Come One [a Buddha]. 13

Not only can the miraculous powers of the Lotus Sūtra grant the upholder the venerative status

and gifts of a Buddha, if a person falls ill, he shall be cured and enjoy eternal youth: ―O

Beflowered by the King of Constellations! With the power of supernatural penetration, you are

to protect this scripture. What is the reason? This scripture, for the people of Jambudvīpa [閻浮

提 Jpn. Enbudai, Ch. Yanfuti], is a good physic for their sickness. If a man has an illness and can

hear this scripture, the illness shall immediately vanish. He shall neither grow old nor die.‖14 It

seems that no matter the ailment, physical or spiritual, the Lotus Sūtra promises salvation,

11
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 175-76. T. no. 262, 9: 34b10-34b12.
12
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 163. T. no. 262, 9: 31b21-b23. Hurvitz notes that to
be cloaked with the Buddha‘s garments means to be protected.
13
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 160. T. no. 262, 9: 30c17-22.
14
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 276. T. no. 262, 9: 54c23-26.

128
because ―[l]ike a clear, cool pond, it can slake the thirst of all. As a chilled person finds fire, as a

naked person finds clothing, as a merchant finds a chief, as a child finds its mother, as a

passenger finds a ship, as a sick person finds a physician, as darkness finds a torch, as a poor

person finds a jewel, as the people find a king, as a commercial traveler finds the sea, as a candle

dispels darkness…‖15 While men ―gain incalculable, limitless merit‖16 upon hearing the twenty-

third chapter, ―The Former Affairs of the Bodhisattva Medicine King,‖17 if a woman ―can accept

and keep it, she shall put an end to her female body, and shall never again receive one‖18—a

significant promise because a woman‘s body was considered polluted and imprisonment in the

female form hindered salvation.

The examples culled here by no means exhaust the extensive promises of the Lotus Sūtra,

but should simply serve to highlight a few of the inherent powers of the scripture. Daniel

Stevenson describes the Lotus Sūtra in the context of practice in China ―as a repository of

religious power and as an object of worship.‖ 19 Tapping into that sacred power ―was usually

articulated in the idiom of stimulus and response. This interactive piety was grounded in concrete

conventions of ritual gesture and devotion, the vocabularies of which were shared across a

diversity of cultic venues—including worship of different sūtras, Buddhas, and bodhisattvas—

and hence not unique to Lotus devotion proper.‖20 The examination of the following sūtras and

their proclamations of power will bear out this assertion, demonstrating the co-constitutive

creation process of miraculous powers manifested.

15
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 274. T. no. 262, 9: 54b14-18.
16
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 275. T. no. 262, 9: 54b27.
17
Jpn. Yakuō bosatsu honji bon, Ch. Yaowang pusa benshi pin; 薬王菩薩本事品.
18
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 275. T. no. 262, 9: 54b27-29.
19
Daniel B. Stevenson, ―Buddhist Practice and the Lotus Sūtra in China,‖ in Readings of the Lotus Sūtra, ed.
Stephen F. Teiser and Jacqueline I. Stone (New York: Columbia University Press, 2009), 146.
20
Ibid., 146.

129
 Much like the Lotus Sūtra,21 the Golden Light Sūtra proclaims itself the ‗king of sūtras

[経王 Jpn. kyōō, Ch. jingwang; Skt. sūtra-rājan].‘ Repeatedly expounding its own excellence,

‗king of sūtras‘ becomes a sort of synonym for the scripture. For example, ―Among purified,

pure, best Bodhisattvas I will preach the excellent Suvar ṇabhāsa, king of sūtras, very profound

on hearing and profound on examination,‖ 22 and ―…it is the king of sūtras, extremely profound,

(and) nothing is found to compare with it. Neither the dust in the Ganges, nor on the earth, nor in

the ocean, nor that found in the sky can provide comparison.‖ 23

And as would be expected from the utmost sūtra, grand pronouncements of power are

frequent. The first chapter of the Golden Light Sūtra catalogues the many woes and distresses

suffered by unenlightened beings:

For those beings whose senses are defective, whose life is expended or failing, beset by
misfortune, their faces averted from the gods, hated by dear, beloved people, oppressed in
such places as households, or at variance with one another, tormented by the destruction
of their property, both in grief and trouble, and in poverty, likewise in the plight of fear,
in the affliction of planet or asterism, in the violent grip of demons, one (who) sees an
evil dream full of grief and trouble…. 24

But relief from these heavy burdens is promised to those who hear the scripture in the proper

religious context, thus activating the vast potency of the sūtra: ―most severe misfortunes are

forever extinguished by the splendour of this sūtra.‖25 By the power of the sūtra, armies of great

and terrifying deities guard those who honor the scripture. Not only do ranks of deities pledge

their protection to those who respectfully hear and uphold the Golden Light Sūtra, the pious are

―honoured throughout numerous millions of aeons by gods, serpents and men, by Kiṃnaras,

21
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 165. See T. no. 262, 9: 32a16.
22
Emmerick, The Sūtra of Golden Light, 1. See T. no. 665, 16: 404a8-10.
23
Emmerick, The Sūtra of Golden Light, 68. T. no. 665, 16: 445a14-17.
24
Emmerick, The Sūtra of Golden Light, 2. T. no. 665, 16: 404a18-24.
25
Emmerick, The Sūtra of Golden Light, 2. T. no. 665, 16: 404a28-29.

130
Asuras and Yakṣas‖ and ―gladly accepted by Buddhas in the ten directions and likewise by the

Bodhisattvas,‖ all the while accumulating ―endless, incalculable, inconceivable‖ merit.26

Vast portions of the sūtra‘s promised rewards are directed at the sovereign, propagating a

hierarchically structured empire with a Buddhist sovereign at the head. The scripture outlines the

virtuous acts of the ideal just king, a cakravartin (wheel-turning) king (天輪聖王; Jpn. tenrin

jōō; Ch. tianlun shengwang), including reproduction and veneration of the Golden Light Sūtra,

adherence to its injunctions and lessons, and a great deal of penitence. In exchange, the power of

the sacred text is unlocked, offering the efficacious protection of the Four Guardian Kings and

ensuring a peaceful, stable country. 27 Even if an ambitious and ruthless king contrives to destroy

the country of a upholder of the Golden Light Sūtra, ―…at that time, at that moment, by the

power of the brilliance of that excellent Suvar ṇabhāsa, king of sūtras, there will arise a conflict

between that neighbouring hostile king and other kings. And there will be regional disturbances

in his own regions. There will be fierce troubles with kings, and diseases caused by planets will

become manifest in his area.‖ 28 As explored in the second chapter, the iconography of the

Chūsonji mandalas proves to be largely concerned with the ideology of the Golden Light Sūtra

and its prescriptions and promises for the cakravartin king.

And seen in the Lotus Sūtra and the Golden Light Sūtra, a similar proclamation of

superiority is issued in the Diamond Sūtra, as the Buddha reveals, ―The Tathagata has taught this

as the highest (paramā) perfection (pāramitā). And what the Tathagata teaches as the highest

perfection, that also the innumerable (aparimāna) Blessed Buddhas do teach. Therefore it is

26
Emmerick, The Sūtra of Golden Light, 2-3. T. no. 665, 16: 404b1-17.
27
Emmerick, The Sūtra of Golden Light, 24-44 and 59-65. T. no. 665, 16: 413c10-417c16 and 427b17-432c10.
28
Emmerick, The Sūtra of Golden Light, 27. T. no. 665, 16: 427c21-23.

131
called the ‗highest perfection.‘‖ 29 And as the ―highest perfection,‖ the Perfection of Wisdom

Sūtras is the mother of all Buddhas:

So fond are the Tathagatas of this perfection of wisdom, so much do they cherish and
protect it. For she is their mother and begetter, showed them this all-knowledge, she
instructed them in the ways of the world. From her have the Tathagatas come forth. For
she has begotten and shown that cognition of the all-knowing, she has shown them the
world for what it really is. The all-knowledge of the Tathagatas has come forth from her.
All the Tathagatas, past, future, and present, win full enlightenment thanks to this
perfection of wisdom. It is in this sense that the perfection of wisdom generates the
Tathagatas, and instructs them in this world. 30

As Edward Conze explains, this all-potent power of the Perfection of Wisdom Sūtras is both the

origin and the outcome of buddhahood.31

Throughout the Perfection of Wisdom Sūtras claims of astonishing power are made with

great frequency. For instance in the third chapter of the Perfection of Wisdom in Eight Thousand

Lines32 in a section titled, ―The Merit Derived from Perfect Wisdom,‖ the scripture claims:

One who will take up this Perfection of Wisdom,

Wherein the Saviours course, and constantly study it;
Fire, poison, sword and water cannot harm him,
And also Mara finds no entrance, nor his host.33

The Buddha, extolling the apotropaic power of the scripture, promises to those who take up the

sūtra that ―[m]en and ghosts alike will be unable to harm them. Nor will they die an untimely

death…. A person who is devoted to this perfection of wisdom will certainly experience no fear,

he will certainly never be stiff with fright—whether he be in a forest, at the foot of a tree, or in

an empty shed, or an open place, or a road, or a highway, or the woods, or on the ocean.‖ 34 The

29
Edward Conze, trans., Buddhist Wisdom: The Diamond Sutra and the Heart Sutra (Toronto: Random House,
2001[1958]), 52.
30
Edward Conze, trans., The Perfection of Wisdom in Eight Thousand Lines and Its Verse Summary, 5th ed. (San
Francisco: Four Seasons Foundation, 1995 [1973]), 172-73.
31
Conze, Buddhist Wisdom, 37.
32
Jpn. Dōgyō hannya kyō; Ch. Daoxing bore jing; Skt. Aṣṭasāhasrikā prajñāpāramitā sūtra; 道行般若経; T. no.
224, 8: 425c3-478b14.
33
Conze, The Perfection of Wisdom, 15.
34
Ibid., 103.

132
power of the Perfection of Wisdom Sūtras cancels karmic debt, releases the pious from the woes

of existence, and even empowers upholders to reach enlightenment.35 Furthermore, as the Heart

Sūtra explains, it is not just people who achieve enlightenment through the salvific power of the

Perfection of Wisdom Sūtras, but ―[a]ll those who appear as Buddhas in the three periods of time

fully awake to the utmost, right, and perfect enlightenment because they have relied on the

perfection of wisdom.‖36

The tremendous merit generated from expounding the Perfection of Wisdom Sūtras for

others, veneration, recitation, and from taking up but one stanza of four lines is explained to

outweigh immeasurably even the most generous of gifts to the Buddhas. In order to demonstrate

the Perfection of Wisdom Sūtras as a repository of power, the Buddha institutes a hierarchy of

various merit-generating gifts that privileges the Perfection of Wisdom Sūtras as the apex. These

comparative metaphors are posed as questions between Subhuti and the Buddha in which gifts

that ―filled this world system of 1,000 million worlds with the seven precious things,‖ 37 ―filled

with the seven precious things as many world systems as there are grains of sand in those Ganges

rivers,‖38 ―renounce all their belongings as many times as there are grains of sand in the river

Ganges,‖39 or ―piled up the seven precious things until their bulk equaled that of all the Sumerus,

kings of mountains, in the world system of 1,000 million worlds,‖ 40 are juxtaposed against the

preferable scenario where ―if a son or daughter of good family had taken from this discourse on

dharma but one stanza of four lines, then they would on the strength of that beget a still greater

35
Conze, Buddhist Wisdom, 55.
36
Ibid., 108.
37
Ibid., 34.
38
Ibid., 47.
39
Ibid., 51.
40
Ibid., 62.

133
heap of merit, immeasurable and incalculable.‖ 41 Again and again, the Buddha extols

engagement with the sūtra as the most salvific path possible.

Again, we find the similar theme of accessing the Buddha through sacred texts iterated in

the Perfection of Wisdom Sūtras. In the Diamond Sūtra, the Buddha assures Subhuti that ―[t]hose

who will take up this discourse on Dharma, bear it in mind, recite, study, and illuminate it in full

detail for others, the Tathagata has known them with his Buddha-cognition, the Tathagata as seen

them with this Buddha-eye, the Tathagata has fully known them.‖ 42 As will be explained later,

such claims are important to understanding the ultimate conflation of sacred texts and the

Buddha—in other words, establishing scripture as dharma relics.

Even this narrow examination of sūtra proclamations reveals the vast potentialities

imbued in sacred text. But how literally were these prescriptions to revere sūtras taken? How

much faith was placed in the power of word, and how do we see this faith manifested? Are the

mandalas reflections of this compelling power? In what follows, I attempt to address these

questions through examinations of various early medieval records that reflect the salvific and

apotropaic power of scripture.

Setsuwa: Wondrous Tales

The efficacious power of sacred word is accessed through proper religious practice, such as

veneration, reproduction, recitation, and dissemination in an interactive relationship between the

text and upholder that allows for the power of the sūtras to be realized. Beyond testifying to the

power vested in sūtra, setsuwa (popular tales often of a religious bent) also reveal the flexible

and open-ended nature of text that makes capable word‘s many and various iterations as

41
Ibid., 47.
42
Ibid., 54.

134
exemplified in the jeweled-stūpa mandalas. As is repeatedly demonstrated in setsuwa, people

beckoned forth this potency embodied in sūtra in myriad ways. Compilations like Priest

Chingen‘s The Great Nation of Japan’s Tales of the Lotus Sūtra (大日本国法華経験記

Dainihonkoku hokekyōkenki) of the mid-eleventh-century record countless examples praising the

redemptive and prophylactic power of sūtras. One account tells the story of two monks who

unwittingly took refuge in a temple that was stalked by a demon.43 The demonic creature,

reeking of cow‘s breath, crashed through the wall of the room where the monks lay sleeping and

dismembered and devoured the older monk. The younger monk, a Lotus Sūtra chanter, clamored

atop the altar, and gripping a statue of Bishamonten recited the Lotus Sūtra throughout the night.

Dawn broke, and the monk discovered the mutilated body of the demon in front of the altar.

Seeing that the spear of Bishamonten was red with the stain of fresh blood, the young monk

realized that the Guardian of the North quelled the evil creature to save a follower of the Lotus

Sūtra. Such is the apotropaic power of the Lotus Sūtra that a pious life will be spared if the

scripture is recited.

Another example relayed through twelfth-century Anthology of Tales from the Past (今昔

物語集 Konjaku monogatarishū) offers the story of an official who, hunted by a voracious

demon previously disguised as a beguiling woman, fled into a cave. When the carnivorous

demon threatens to continue his pursuit of the official, a disembodied voice sounds from within

the cave, commanding the demon to retreat. The voice is revealed to be the first character of the

Lotus Sūtra, myō 妙, the last but potent remaining part of a wind-battered copy of the scripture

43
Yoshiko Kurata Dykstra, trans., Miraculous Tales of the Lotus Sutra from Ancient Japan: The Dainihonkoku
Hokekyōkenki of Priest Chingen (Osaka: Intercultural Research Institute, Kansai University of Foreign Studies,
1983), 82-83.

135
once enshrined within a now fallen stūpa. 44 The setsuwa admonishes the audience to realize that

―though only one character of the Lotus Sūtra remained, it saved a man‘s life. You can imagine,

then, the merit that will come from copying the Lotus Sūtra in the prescribed form and with true

faith. If such is the benefit in this present life, do not doubt that you will escape all torments in

the life to come.‖45

Often setsuwa testify to the miraculous healing power of sūtra. The Miraculous Episodes

of Good and Evil Karmic Effects in the Nation of Japan (日本国現報善悪霊異記 Nihonkoku

genpō zen'aku ryōiki) records just such an instance. A respected monk, Chōgi 長義, loses the

sight in one of his eyes without explanation. Distressed and ashamed at his misfortune, he

gathers many monks to recite for three days and nights the Diamond Sūtra. Amazingly, the

monk‘s eyesight is returned, and the setsuwa proclaims, ―How great is the miraculous power of

the Hannya! For, if a vow is made with profound faith, it will never remain unfulfilled.‖ 46

Piously copied sūtras even have the extraordinary ability to transform into flesh, a

phenomenon Charlotte Eubanks describes as ―text made flesh.‖ 47 Some stories claim that even

reciting in a mocking fashion the title of the Lotus Sūtra is enough to spare a sinner from a

tortuous hell, as was the case of an unbeliever named Sonko.48 After Sonko mocked and

ridiculed a Lotus devotee, causing the pious man to drop the copy of the Lotus Sūtra he wore

around his neck, Sonko collected the sūtra and took it home, only to forget about it. Years later,

44
Marian Ury, Tales of Times Now Past: Sixty-Two Stories from a Medieval Japanese Collection (Ann Arbor:
Center for Japanese Studies, University of Michigan, 1993), 87-89.
45
Ibid., 89.
46
Kyoko Motomochi Nakamura, trans., Miraculous Stories from the Japanese Buddhist Tradition The Nihon Ryōiki
of the Monk Kyōkai (Cambridge, MA: Harvard University Press, 1973), 249. Keikai 景戒, ―Nihon Ryōiki 日本霊異
記,‖ in Nihon koten bungaku zenshū 日本古典文学全集, ed. Nakada Norio 中田祝夫, vol. 6 (Tokyo: Shōgakkan,
1975), 311-12.
47
Eubanks, ―Rendering the Body Buddhist: Sermonizing in Medieval Japan,‖ 314.
48
The story comes from the twelfth-century Summary Notes of One Hundred Lectures on Dharma (百座法談聞書
抄 Hyakuza hōdan kikigakishō). See Satō Akio 佐藤亮雄, ed., Hyakuza hōdan kikigakishō 百座法談聞書抄
(Tokyo: Nan‘undō Ofūsha, 1963), 133-35.

136
he died and faced the judgment of King Enma (閻魔王 Ch. Yanluowang; Skt. Yama Rājā). As he

was about to be sentenced, a kindly demon in attendance reminded Enma that Sonko had recited

the title of the Lotus Sūtra, albeit sarcastically and cruelly. This one recitation was enough to

send Sonko back to life. The demon then told the man that he was an incarnation of the last

remaining Lotus scroll of the tattered sūtra left exposed and forgotten in the man‘s house. 49 Such

testimonials recounting the efficacious power of the Lotus Sūtra abound, suggesting the

prevalent belief in the scripture as a talisman and active agent capable of transformations

involved in one‘s personal salvation, be it from imminent physical danger or from eternal

damnation.

Another such story comes from the Miraculous Episodes of Good and Evil Karmic

Effects in the Nation of Japan in which a devoted reciter of the Heart Sūtra and copier of other

scriptures was summoned to the court of King Enma after her death (painlessly we are assured)

so that she might chant sūtras before him, allowing him to witness and revel in the beauty of her

celebrated voice. 50 After three days, she is allowed to return to life. She then notices three men in

yellow robes standing by the gate who explain to her that this encounter is not their first and that

at the Nara east market in three days time, they will meet again. It is at the market that the

woman purchases two scrolls of the Brahma Net Sūtra51 and one scroll of the Heart Sūtra and

afterward realizes that these scriptures are in fact her own copies made years before on yellow

paper. Furthermore, she discovers the sūtras to be none other than the three men of yellow robes.

The inherent power of the Buddha‘s words to not only protect, guide, and comfort; but also to

perform miraculous transformations surely propelled the continuously (re)created lives of texts.

49
Satō, Hyakuza hōdan kikigakishō, 134-35.
50
Nakamura, Miraculous Stories from the Japanese Buddhist Tradition, 186-87. Keikai, ―Nihon Ryōiki,‖ 197-99.
51
Jpn. Bonmōkyō; Ch. Fanwangjing; Skt. Brahmajāla sūtra; 梵網経; T. no. 1484, 24: 997a-1010a.

137
 Collections of setsuwa often record surikuyō 摺供養, a ritual in which sūtras are copied

and the merit dedicated to deceased loved ones so that they might improve their karmic lot in the

form of a more advantageous rebirth or at least to lessen the anguish and physical torment

inflicted in hell. For example, the Miraculous Episodes of Good and Evil Karmic Effects in the

Nation of Japan describes an occasion when King Yama (another name for King Enma)

summoned Fujiwara no Asomi Hirotari 藤原朝臣広足 to hell at the request of his suffering wife.

Having already endured three years of her six-year punishment, she wishes for her husband to

shoulder some of the burden since her death was caused by childbirth. Hirotari promises to return

to the world and copy, expound, and recite the Lotus Sūtra in order to dedicate the merit to his

suffering wife.52

Examples within setsuwa manifesting the extraordinary powers of sacred word, while

varied and fascinating, are too numerous to discuss in detail here.53 The episodes presented here

have been chosen because they are representative of the wide spectrum of efficacious powers

believed to reside in scripture—from defense against demonic attacks and spontaneous healing to

salvation from hell and relief from tortuous suffering. Miraculous tales from China, such as those

documented in the Accounts in Dissemination and Praise of the Lotus [Sūtra] (弘贊法華傳

Hongzan fahua zhuan) record feats, equally as astonishing, performed or made possible by

sūtra‘s power.

These ubiquitous tales of extraordinary performances manifested through scripture‘s

inherent power suggest the open-ended nature of text that makes possible scripture‘s numerous

and diverse iterations. The limitless potential of vivified sacred word not only to generate merit

52
Nakamura, Miraculous Stories from the Japanese Buddhist Tradition, 233-35. Keikai, ―Nihon Ryōiki,‖ 281-84.
53
For recent works on the subject of setsuwa, see Eubanks, ―Rendering the Body Buddhist;‖ and Howell, ―Setsuwa,
Knowledge, and the Culture of Reading and Writing in Medieval Japan.‖

138
but also to transform and act on the world encourages text to manifest in various ways within

Buddhist visual culture and religious practice. This active and flexible nature of scripture is

revealed in the textuality of early medieval society as reverent objects in the forms of relic

deposits, elaborate and exquisite sūtra scrolls—sometimes incorporating bodily offerings such as

blood or hair, layered images of sacred text and mundane picture, and even sūtra as relic

constructing reliquary in the jeweled-stūpa mandalas.

Commenting on the Power of Sacred Word

Examining ecclesiastical commentaries reporting on sacred texts‘ significance to Buddhism,

society, and the cosmos helps establish the place of sūtra in early medieval Japan and its role and

potentialities in visual culture. Much like the fluid relations between the porous systems of

Buddhist thought in early medieval Japan, the jeweled-stūpa mandalas cannot be pigeonholed

into a particular school. Therefore, in this section on monastic commentaries, I analyze the

general commendations and concerns reflected in the writings of several early medieval monks

associated with different schools, exposing the often shared understandings of the active, salvific,

and foundational nature of sacred text as well as its limits and dangers if carelessly regarded or

invoked. However, this is not to imply a universal concept of the role of sacred text in Buddhism.

Texts and the exercise of writing certainly did not have the same meaning for all, and therefore I

wish to avoid a homogenized characterization, but rather to present some examples of the

textualized context out of which the innovative jeweled-stūpa mandalas emerged.

As one of the most prolific writers on the power of language and sacred text, Kūkai is a

good place to start when considering the role of word in early medieval Buddhist Japan. As

Ryūichi Abé demonstrates throughout his book, Kūkai‘s writings accomplished a great deal in

139
the shifting attitudes toward language and text.54 It is because of this that I devote more space to

his ideas than other commentators, although this cursory introduction to Kūkai‘s theories on the

power and potential of language is in no way comprehensive. 55 The revolutionary approach to

text promoted by Kūkai represented a drastic break from the general considerations of language

in the eighth century. He re-characterized56 the very nature of language and its origins, asserting

in the Shōji jissōgi 声字実相義57 that the Sanskrit letter A is the Dharmakāya‘s seed mantra and

thus is the progenitor of all letters, words, languages, and indeed all things as the ‗originally

nonarising‘ (本不生 Jpn. honpushō, Ch. benbusheng; Skt. ādyanutpāda). He explains:

It is the wheel of letters or the syllabary given in the Vajraśekhara Sūtra [金剛頂経 Jpn.
Kongōchōkyō, Ch. Jingangjing] and the Mahāvairocana Sūtra [大日経 Jpn. Dainichikyō,
Ch. Darijing]. By the syllabary is meant A, Ha, etc. in the Sanskrit alphabet. A, etc. are
the namewords, secret designations, of the Dharmakāya Tathāgata. Gods, serpents [Skt.
nāgas 竜 Jpn. ryū, Ch. long], demons, etc. also have their respective syllabary. Yet the
root of them is in the fountainhead of [the king of mantras of] Mahāvairocana
[Dharmakāya]. Emanated from this and ramified on and on are the languages current in
the world. If a man knows the true significance of this, we call him on who knows the
true words [mantra]. If he does not know the fountainhead, we call him one who uses
false words. The use of false words makes one subject to sufferings in long nights of
darkness. The differences are precisely those between medicine and poison,
enlightenment and delusion, or gain and loss.58

By declaring the Sanskrit syllabary, and indeed all languages, to be identical with the

Dharmakāya; and by declaring that all language is mantra, an image of the universe as cosmic

text is articulated. Kūkai identifies the ten realms 59 as matters of semiotic differentiation (十種文

字 jusshu monji). Using a vertical reading of the world scheme, the highest and most perfect

54
Abé, The Weaving of Mantra.
55
For a nuanced explication, see Abé, The Weaving of Mantra.
56
For the Indic and Chinese origins of Kūkai‘s concepts of language and text, see Abé, The Weaving of Mantra; and
Payne, ―Awakening and Language,‖ 79-96.
57
Yoshito S. Hakeda, trans., Kūkai: Major Works (New York, Columbia University Press, 1972), 234-246. Abé
translates the title of this treatise, Voice, Letter, Reality, whereas Hakeda chooses to translate it as The Meanings of
Sound, Word, and Reality.
58
Hakeda, Kūkai, 242.
59
The ten realms are as follows: the realms of hell, hungry ghosts, animals, asuras, humans, heaven, śrāvakas,
pratyekabuddhas, bodhisattvas, and Buddhas.

140
language is that of the Buddhas (mantra). However, by horizontally interpreting the realms, the

languages of all ten spheres are none other than mantra as he explains: ―All sorts of names

(signs) originate from the Dharmakāya. They all issue forth from it (him) and become the

languages circulating in the world. The language that is aware of this truth is called the true word

(shingon) and other languages that are not conscious of their source are called illusory words

([妄語] mōgo).‖60

Summarizing a key point in the Ten Abiding Stages of Mind According to the Secret

Mandalas (秘密曼荼羅十住心論 Himitsu mandara jūjūshinron),61 Ryūichi Abé describes

Kūkai‘s conclusion as, ―the universe itself, as it is, is the Dharmakāya‘s body made up of the

sacred letters, the body of the text manifesting itself as the realm of the ultimate reality, his

palace.‖62 This radical concept of language as originating in the dharmakāya institutes a vision of

the world as textual imbrication: everything is text, and thus text constructs everything and is the

root of all things. There exists nothing that is not encapsulated by sacred text, nothing that does

not issue forth from it, for differentiation is a matter of semiotic articulation and signification (差

別 shabetsu);63 in essence, language produced the universe, and thus all is world-text. This

revelatory claim leads Fabio Rambelli to assert the ultimate value of texts as not just signs but

―microcosms, holographs of the dharma-realm.‖64 Kūkai has removed language and text from the

mundane world of humans and revealed it to be the embodiment of the dharmakāya, and thus

60
Abé, The Weaving of Mantra, 283; and Hakeda, Kūkai, 240-41.
61
Throughout his writings on text, body, and dharmakāya, Kūkai refers back to the Avataṃsaka Sūtra‘s
characterization of the world as scripture text. See Abé, The Weaving of Mantra; and Gómez, ―The Whole Universe
as a Sūtra,‖ 107-112.
62
Abé, The Weaving of Mantra, 334.
63
For a thorough analysis of Kūkai‘s argument of language as differentiation, see Abé, The Weaving of Mantra,
275-304.
64
Rambelli, ―Texts, Talismans, and Jewels,‖ 73

141
emptiness (空 Jpn. kū, Ch. kong; Skt. śūnyatā). Sūtras therefore contain all things of the world,

and the world reflects back as sūtra.

Kūkai‘s theories on language advocate a positive evaluation of the possibilities of

language in the process of enlightenment or awakening. By holding that Dharmakāya

Mahāvairocana (大日如來 Jpn. Dainichi, Ch. Dari Rulai) is constantly preaching the dharma

without cessation (法身説法 Jpn. hōsshin seppō, Ch. fashen shuofa), the argument is advanced

that enlightenment is possible through language. 65 This recasting of language contradicts other

Buddhist schools, which limit the power and abilities of language, maintaining that ―language

cannot express, and ordinary dualistic cognition cannot grasp, the reality of emptiness and

interdependence.‖66 And while Kūkai himself lamented the limits of language, mostly in the

context of exoteric teachings, 67 he nonetheless advocated a strong position for the ultimate value

of text.68

Important for discussions of material manifestations of scripture is Kūkai‘s notion of

open-text. Kūkai‘s all-inclusive theories on text demand that text remain an open and active

manuscript, because while text reflects all of universe, it is a scripture in flux, fluid and dynamic.

Such a concept encourages diverse visualizations of sacred word. Jacques Derrida espouses a

similar view of text, arguing that no text is purely self-referential or closed, a subject I return to

in the sixth chapter.

65
For a discussion on this subject, see Abé, The Weaving of Mantra; David Gardiner, ―Kūkai's View of Exoteric
Buddhism in his Benkenmitsu nikyōron,‖ Bulletin of the Research Institute of Esoteric Buddhist Culture 5 (1992):
161-202; Payne, ―Awakening and Language,‖ 79-96; and Fabio Rambelli, ―The Semiotic Articulation of Hosshin
Seppō: An Interpretive Study of the Concepts of Mon and Monji in Kūkai's Mikkyō,‖ in Esoteric Buddhism in
Japan, ed. Ian Astley (Copenhagen and Aarhus: The Seminar for Buddhist Studies, 1994), 17-36.
66
Payne, ―Awakening and Language,‖ 90.
67
Distinguishing the Two Teachings (Benkenmitsu nikyōron 弁顕密二教論). Hakeda, Kūkai, 151-57.
68
For a discussion on Kūkai‘s prescriptions for textual use and accessing and harnessing the power inherent in
sacred word, see in particular Abé, The Weaving of Mantra, ―Part I: Origin, Traces, Nonorigin,‖ 67-184 and ―Part
III: Writing and Polity,‖ 273-398.

142
 Shingon Buddhism is not the only school of Buddhist thought in early medieval Japan to

have advocated a prominent and positive role of language. Many associated with Tendai and

Nichiren promoted a particularly Lotus Sūtra-centric view of worship and enlightenment. Greatly

impacted by the writings of Zhiyi, Saichō repeatedly praised the Lotus Sūtra as the ultimate

scripture of truth and the path to the final awakening 69 and ordered that the Lotus Sūtra be

preached at all times in the samādhi hall (三昧堂 Jpn. sanmaidō, Ch.sanmeitang) on Mt. Hiei. 70

Indeed, a crucial stage along the path toward awakening in original enlightenment discourse (本

覚 Jpn. hongaku, Ch. benjue) is ‗verbal identity‘ (名字即 Jpn. myōjisoku, Ch. mingziji).71 The

verbal identity stage in original enlightenment thought is the moment when one realizes through

pious interaction with the words of the sūtras, either through reading or hearing an explication,

that all things are in fact identical with the buddhadharma. As Jacqueline Stone surmises, ―From

this perspective, there could be no enlightenment unmediated by words; only by reading the

characters of the sūtra or hearing an explication of doctrine could original enlightenment be

realized.‖72 The integrality of words to enlightenment privileges sūtras for their inherent salvific

power. The Digest of the Light of Han (漢光類聚 Kankō ruijū),73 a thirteenth-century Tendai

text of oral transmissions, claims that ―written words are not [merely] written words; language is

69
For a thorough analysis of Saichō and his impact, see Paul Groner, Saichō The Establishment of the Japanese
Tendai School (Honolulu: University of Hawai‘i Press, 2000). For an analysis of original enlightenment during
medieval Japan, and in particular its relationship to Tendai and Nichiren thought, see Jacqueline I. Stone, ―Medieval
Tendai Hongaku Thought and the New Kamakura Buddhism: A Reconsideration,‖ JJRS 22 (1995): 17-48, and
Stone, Original Enlightenment and the Transformation of Medieval Japanese Buddhism.
70
Howell, ―Setsuwa, Knowledge, and the Cultures of Reading and Writing in Medieval Japan,‖ 179.
71
See Groner, Saichō; and Stone, Original Enlightenment and the Transformation of Medieval Japanese Buddhism.
And for a discussion in Tiantai Buddhism, see Neal Donner and Daniel B. Stevenson, The Great Calming and
Contemplation: A Study and Annotated Translation of Chih-i’s Mo-ho Chih-Kuan (Honolulu: University of Hawai‘i
Press, 1993), 207-18.
72
Jacqueline I. Stone, ―‗Not Mere Written Words‘: Perspectives on the Language of the Lotus Sūtra in Medieval
Japan,‖ in Discourse and Ideology in Medieval Japanese Buddhism, ed. Richard R. Payne and Taigen Dan Leighton
(London and New York: Routledge, 2006), 162.
73
See Groner, Saichō; and Stone, Original Enlightenment.

143
liberation‖ in response to the criticisms of extreme attachment to—yet denial of the power of—

written texts. 74 The Digest of the Light of Han explains that

Each word and phrase is in every case endowed with the eight aspects [of the Buddha‘s
career]. Thus we speak of the principle that written words are precisely liberation.
Ignorant persons do not know this meaning, and so they either cling to words and letters,
or reject words and letters altogether. Neither way will do….The Denbōketsu 75 states,
‗The Great Teacher Nanyue [Huisi, 515-577] said, ‗Words are none other than liberation.
If one seeks liberation apart from words, there is no such place [where it can be found].‘ 76

Nichiren fervently promoted the Lotus Sūtra as the supreme Buddhist authority

subsuming all other doctrines and praxis. 77 He advocated chanting the sūtra‘s title as the mantra:

namu myōhō rengekyō 南無妙法蓮華経 (homage to the Lotus Sūtra). According to Nichiren, the

power to realize buddhahood in this very body was contained in the characters of the title. 78 Thus,

through the mobilization of language, the title of the Lotus Sūtra is the key to religious practice

and salvation. Based on this advocacy of the inherent power within the characters of the Lotus

Sūtra‘s title, Nichiren created the Great Mandala 大曼荼羅. This calligraphic mandala, brushed

first by Nichiren and later by his followers, championed the power invested in word.79

But in monastic commentaries, texts were commonly treated in a highly prescriptive

manner. Kūkai, Saichō, Kakuban, and Genshin, to mention a few, all prescribed very specific

directions for accessing the power in sacred texts and warned against improper use, citing dire

consequences for uninformed or reckless handling. Much of the idea of limiting production of

and access to texts originated in Confucian attitudes toward language and writing which had as

its vehicle the ritsuryō 律令 system of government, which standardized administrative and penal

74
Ibid., 174.
75
Stone clarifies that this text is likely a reference to the oral transmission Saichō received in China. Stone, ―‗Not
Mere Written Words,‘‖ 189 n22.
76
Stone, ―‗Not Mere Written Words,‘‖ 168.
77
Stone, Original Enlightenment, 261.
78
Ibid., 241.
79
For more on this topic, see chapter six.

144
codes under a centralized state with the emperor as its head. The strict control over sacred texts

and the high value placed on literacy during the Nara and early- through mid-Heian periods

suggests the power to be gained by possessing and composing texts. A nuanced and detailed

discussion of the role of early medieval sacred text as illustrated through ecclesiastical

commentaries is well beyond the limits of this study. What I have tried to do in this section is to

introduce and highlight some of the prominent, influential, and revolutionizing concepts of

sacred language.

Dharma Relics

Any examination of dharma relics must include a discussion of the notion of dharmakāya;

however, beyond the necessary references to the concept, I reserve the analysis of the bodies of

the Buddha for the following chapter. Instead, I concentrate on the nonduality of the Buddha and

the sūtras, which establishes the theoretical underpinning the discussion of dharmakāya

elaborated in chapter five. I offer this equivalence as the basis for the category of relics known as

dharma relics. By briefly exploring this nonduality, we might understand the origin and basis of

the power invested in sūtras and the impetus compelling the ubiquitous and diverse visual

manifestations of this power.

It should be noted that dharma does not have merely one definition, but instead is a

multilayered idea of interrelated concepts. Based on the commentaries by Dignāga and

Vasubandhu on the Perfection of Wisdom Sūtras, dharma can be ―thought of as nondual

awareness, as a book or collection of teachings, as a path, or as the cessation of suffering.‖ 80

Asaṅga‘s commentary categorizes dharma as the collection of teachings—which can be

80
Malcolm David Eckel, To See the Buddha: A Philosopher's Quest for the Meaning of Emptiness (Princeton:
Princeton University Press, 1994), 100.

145
materially manifested in the form of sūtras; dharma as understanding—meaning both the goal

(buddhahood, Skt. buddhatva) and the path to cessation; and dharma as nondual awareness. 81

And as it is frequently described in canonical texts, dharma is ―beautiful in the beginning,

beautiful in the middle, and beautiful in the end.‖ 82

Even in early texts we find evidence of the nonduality of the Buddha and the dharma, or

his teachings. In the non-canonical Pali text, Sūtra on the Questions of King Miliṇḍa,83 the

Buddha declares that one who sees the dharma thus sees the Buddha. 84 A similar sentiment in the

Saṃyutta Nikāya of the Pali canon discloses a conversation between the Buddha and a sickly

Vakkali who longs to see the Tathāgata. Gautama exposes the error of Vakkali‘s desire saying,

―What is there in seeing this vile body (pūti-kāya) of mine?85 He who seeth the Norm

[dhamma/dharma], Vakkali, he seeth me: he who seeth me, Vakkali, he seeth the Norm. Verily,

seeing the Norm, Vakkali, one sees me: seeing me, one sees the Norm.‖ 86 The distinction drawn

here is one of corruptibility versus the true essence of the Buddha-nature encapsulated in his

teachings. Elsewhere in the Saṃyutta Nikāya, the juxtaposition is reiterated, although this time

the Buddha explains the inevitable decay of the human body: ―[t]his body be devoured by crows

and vultures, devoured by kites and dogs.‖87 In the Perfection of Wisdom in Eight Thousand

Lines, Dharmodgata 曇無竭 explains the fool‘s errand: ―Equally foolish are all those who adhere

81
Ibid., 101. For a good introduction to the many understandings of dharma/dhamma, see The Pali Text Society’s
Pali-English Dictionary‘s extensive entry for dhamma. T.W. Rhys Davids and William Stede, eds., The Pali Text
Society’s Pali-English Dictionary (Chipstead, Surrey: The Pali Text Society, 1925).
82
The Pali Text Society’s Pali-English Dictionary, entry on dhamma, which lists some of the many occurrences.
83
Jpn. Nasen biku kyō; Ch. Naxian biqiu jing; Skt. Miliṇḍapañha; 那先比丘経; T. no. 1670a, 32: 694a4-719a20.
84
Mrs. Rhys Davids, The Milinda-Questions: An Inquiry into Its Place in the History of Buddhism with a Theory as
to Its Author (London: Routledge, 2000 [1930]), 110. Davids‘ exact translation reads as thus: ―Just so, great king,
whosoever sees what the Truth [dharma] is, he sees what the Blessed One was, for the Truth was preached by the
Blessed One.‖
85
According to The Pali Text Society’s Pali-English Dictionary, entry for kāya, pūti-kāya refers to the foul body or
physical body of the Buddha,which is finite.
86
F.L. Woodward, trans., The Book of Kindred Sayings (Sanyutta-Nikāya) or Grouped Suttas, vol. 3 (London:
Oxford University Press, 1917-30), 102-03.
87
Woodward, The Book of Kindred Sayings (Sanyutta-Nikāya) or Grouped Suttas, vol. 5, 320-21.

146
to the Tathāgata through form and sound, and who in consequence imagine the coming or going

of a Tathāgata. For a Tathāgata cannot be seen from his form-body. The Dharma-bodies are the

Tathāgatas and the real nature of dharmas does not come or go.‖88 Again, such comparisons cast

the true Buddha-essence as embodied only in the dharma. The Diamond Sūtra also confirms this

nonduality and its significance as the path to enlightenment:

Those who by my form did see me,

And those who followed me by voice
Wrong the efforts they engaged in,
Me those people will not see.
From the Dharma should one see the Buddhas,
From the Dharmabodies [dharmakāya] comes their guidance.
Yet Dharma‘s true nature [dharmatā] cannot be discerned,
And no one can be conscious of it as an object. 89

As further evidence of this nonduality, the Perfection of Wisdom Sūtras stake claim as the

genetrix of the Buddhas, as discussed the previous section.

The oft-invoked dialogue from the Nirvāṇa Sūtra90 between Śākyamuni and Ānanda

around the time of the parinirvāṇa further demonstrates the identity of the Buddha with the

dharma and the authority and power invested in the dharma:

From the beginning, Ānanda, I have taught you that whatever things are delightful and
desirable, joyful and pleasing, these are subject to separation and destruction, to
disintegration and dissociation. So Ānanda, whether now or after my decease, whoever
you are, you must remain as islands to yourselves, as defences to yourselves with the
Dharma as your island and the Dharma as your defence, remaining unconcerned with
other islands and other defences. If you ask the reason for this, then know that whether
now or after my decease, whoever remain as islands to themselves, as defences to
themselves, with the Dharma as their island and the Dharma as their defence, not
concerning themselves with other islands and other defences, such ones are the foremost
of my questing disciples. 91

88
Conze, The Perfection of Wisdom in Eight Thousand Lines and Its Verse Summary, 291.
89
Conze, Buddhist Wisdom, 63.
90
Jpn. Dai nehan kyō; Ch. Da banniepan jing; Skt. Mahāparinirvāṇa sūtra; 大般涅槃経; T. no. 374, 12: 365c-603c.
91
David L. Snellgrove, ―Śākyamuni's Final ‗nirvāṇa,‘‖ Bulletin of the School of Oriental and African Studies 36, no.
2 (1973): 399-411.

147
 The Perfection of Wisdom in Eight Thousand Lines offers a perspective on relics: ―O

Kauśika, the Tathāgata attains his body (śarīra) through the skill-in-means of the Perfection of

Wisdom. This [body] is the location (āśraya) of omniscience. At this location omniscience

comes into being, the Buddha relic (śarīra) comes into being, the Dharma relic (śarīra) comes

into being, and the Saṃgha relic (śarīra) comes into being.‖92 Designating textual dharma as

relic is also found in the Lotus Sūtra, and within this scripture, the equivalence of the Buddha

and the dharma, and not surprisingly, the Lotus Sūtra as the ultimate or true vehicle of the law is

articulated. ―O Medicine King! Wherever it may be preached, or read, or recited, or written, or

whatever place a roll of this scripture may occupy, in all those places one is to erect a stūpa of

the seven jewels, building it high and wide with impressive decoration. There is no need even to

lodge śarīra in it, what is the reason? Within it there is already a whole body of the Thus Come

One.‖93 Again, the scripture equates the sūtra with the Buddha, saying, ―If there is anyone who

can hold [the Lotus Sūtra], / Then he holds the Buddha body‖ 94 and ―if there is a man…who

shall look with veneration on a roll of this scripture as if it were the Buddha himself….‖ 95 The

Lotus Sūtra and Its Traditions (法華伝記 Jpn. Hokke denki, Ch. Fahua zhuan ji), an eighth-

century text expounding the glories and benefits of the Lotus Sūtra written by the Chinese monk,

Sengxiang 詳撰, proclaims that each character of the Lotus Sūtra is a Buddha.96 The nonduality

of the sūtra and the body of the Buddha as scriptural text represents the ultimate conflation of

dharma and relic, and thus constitutes the dharma relic category of relic veneration. It is the

understanding that the dharma preached by the Buddhas is in essence the dharmakāya—which

92
Eckel, To See the Buddha, 98. The ambiguities surrounding the terms for relic and body and the resulting
conceptual possibilities are discussed in the next chapter. Haribhadra‘s commentary on the sūtra also classifies the
scripture as dharma relics.
93
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 163. T. no. 262, 9: 31b26-29.
94
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 176. T. no. 262, 9: 34b12.
95
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 159. T. no. 262, 9: 30c11-13.
96
Kuno and Nakamura, Bukkyō bijutsu jiten, 842.

148
was early on specifically understood as the concrete, material doctrines captured in sūtras—that

so imbues sacred word with authoritative power and soteriological sway.

Dharma-verse relics and dhāraṇ īs constitute receptacles impregnated with great power

believed to capture in a condensed and potent form the whole of the dharma. For instance, the

pratītyasamutpādagāthā, commonly referred to as the ‗Buddhist creed‘ or the ye dharmā

hetuprabhavā verse, was honored as a powerful distillation of the Buddha-essence as

encapsulated by the doctrine of the ‗dependent origination‘ (縁起 Jpn. engi, Ch. yuanqi; Skt.

pratītyasamutpāda) and often enshrined as a relic of the Buddha. As Daniel Boucher has

demonstrated, this verse is revealed as nondual with the Buddha‘s dharma in several sūtras. For

instance, the Śālistamba Sūtra, a canonical text on the pratītyasamutpāda, declares, ―He, monks,

who sees the pratītyasamutpāda sees the dharma; he who sees the dharma sees the Buddha.‖ 97

The Sūtra on the Merit of Bathing the Buddha,98 supposedly translated by Yijing, also reveres

the pratītyasamutpādagāthā as the dharma relic of the Buddha:

After my nirvāṇa, if you wish to do homage to these three bodies [dharmakāya,

saṃbhogakāya, and nirmāṇakāya], then you should do homage to my relics. But there are
two kinds: the first is the bodily relic; the second is the dharma-verse relic. I will now
recite the verse:

All things arise from a cause.

The Tathāgata has explained their cause
And the cessation of the cause of these things.
This the great ascetic has explained. 99

And the Pratītyasamutpāda Sūtra, a short Mahāyāna text, also reveals the verse to be a dharma

relic worthy of veneration through enshrinement. 100

97
Daniel Boucher, ―The Pratityasamutpadagatha and Its Role in the Medieval Cult of the Relics,‖ The Journal of
the International Association of Buddhist Studies 14 (1993): 2.
98
Jpn. Yokubutsu kudoku kyō; Ch. Yufo gongde jing; 浴佛功德経; T. no. 698, 16:799c4-800c15.
99
Daniel Boucher, ―Sūtra on the Merit of Bathing the Buddha,‖ in Buddhism in Practice, ed. Donald S. Lopez
(Princeton: Princeton University Press, 1995), 65.

149
 Dhāraṇ īs, while most often associated with esoteric rites, actually make frequent

appearances in exoteric texts. Indeed, the Perfection of Wisdom Sūtras, the Lotus Sūtra, and the

Golden Light Sūtra all contain dhāraṇīs loaded with such promises as protection from harm and

miraculous recovery from illness or disease. But as Abé notes, they function more as

―appendages to the sūtras‘ main body of text. As a linguistic device for accelerating the learning

process, dhāraṇī recitation is auxiliary to the reading, understanding, and memorizing of a

sutra.‖101 On the other hand, esoteric dhāraṇ īs require quite different semiological actions from

the encounterer of the text, the ritual participant, and even from the sūtra itself: ―It is no longer

the reading, reciting, and memorizing of the sūtra but the ritual actions prescribed in the sūtra

that provide the context for recitation of the dhāraṇī. That is to say, the esoteric sūtra partakes of

the function of a ritual manual.‖ 102 And much like the self-ascribed powers of sūtras, dhāraṇīs

can lay claim to considerable potency perceived to have emanated from the dhāraṇī‘s root as an

expression of the essence of the dharmakāya and thus the embodiment of wisdom.

Ritualistic Constitution of Dharma Relics

In the examination of the potency of sacred word, certain questions beg consideration. At the

heart of the matter lies the issue of when sūtra text becomes a dharma relic. Is there a time when

scripture is in fact not a relic? At what point does the transubstantiation of ordinary into sacred

language occur? Or as Stanley Tambiah has posed, ―If sacred words are thought to possess a

special kind of power not normally associated with ordinary language, to what extent is this due

100
See Richard Salomon and Gregory Schopen, ―The Indravarman (Avaca) Casket Inscription Reconsidered:
Further Evidence for Canonical Passages in Buddhist Inscriptions,‖ The Journal of the International Association of
Buddhist Studies 7, no. 1 (1984): 117, which provides a translation from the Tibetan text: ―If a devoted son or
daughter of good family were to make on an unestablished place (apratiṣṭhite deśe or pradeśe) a stūpa the size of an
āmalaka fruit—with a yaṣṭi the size of a needle and an umbrella the size of a bakula flower—and were to put in it
the verse of the Dharma-relic of pratītyasamutpāda, he would generate brahmic merit (brāhmapuṇyaṃ prasavet).‖
101
Abé, The Weaving of Mantra, 166-67.
102
Ibid., 167.

150
to the fact that the sacred language as such may be exclusive and different from the secular or

profane language?‖103 Are the modern renditions of the sūtras we hold in our hands in scholastic

pursuit, in fact, relics? What role did the invention and common incorporation of printed versions

of scripture have on notions of dharma relics? The overall question is not unlike the activation or

vivification of icons; it is often the ritualistic placement of relics, including dharma relics, inside

a sculpture that transforms what was merely mundane material of wood, metal, and clay into an

icon partaking in the essence of the Buddha. 104

I proffer that the ways in which textual dharma came to be conceived of as relics of the

Buddha was not only through direct statements as such and through the nondual conflation at the

fundamental level of the Buddha-essence, or ultimate truth and the dharma as understanding and

dharma as text, but perhaps more importantly through the treatment of sacred text as relic in

religious practice. As William Graham has astutely observed,

A text becomes ‗scripture‘ in active, subjective relationship to persons, and as part of a

cumulative communal tradition. No text, written or oral or both, is sacred or authoritative
in isolation from a community….A book is only ‗scripture‘ insofar as a group of persons
perceive it to be sacred or holy, powerful and portentous, possessed of an exalted
authority, and in some fashion transcendent of, and hence distinct from, all other speech
and writing. 105

Without the appropriate context, dharma relics are otherwise used text, read, scribbled upon,

dissected, or untouched—words devoid of meaning. Because scriptures–as repositories of great

and sacred power and as dharma relics of the Buddha and manifestations of the dharmakāya

(unlike relics of corporeality)–are at the same time the only records of the Buddhas‘ teachings

and instructions, by necessity they must straddle the line demarcating the sacred from the

103
Stanley Jeyaraja Tambiah, Culture, Thought, and Social Action: An Anthropological Perspective (Cambridge:
Harvard University Press, 1985), 22.
104
Another common ritualistic method of icon vivification is in the painting of the eyes.
105
William Graham, Beyond Written Word: Oral Aspects of Scripture in the History of Religion (Cambridge:
Cambridge University Press, 1993), 5.

151
practical. It is through the proper veneration of the sūtras as sacred, empowered objects that the

transubstantiation of paper into relic occurs. Through ritualistic preparation, veneration, and

visual cues, such as in the elaborate scripture transcription in the jeweled-stūpa mandalas, sūtra is

revealed to be sacred and powerful embodiments of the Buddha.

Scriptures themselves suggest that sūtras should be venerated as one would a Buddha, or

for that matter, an icon.106 At multiple points, the Perfection of Wisdom Sūtras are worshiped by

devotees ―with flowers which they had brought along, and with garlands, wreaths, raiment,

jewels, incense, flags and golden and silvery flowers, and one after another, they deposited their

portion in front of it‖107 just as you would a Buddha with ―heavenly flowers, incense, perfumes,

wreaths, ointments, aromatic powders, jewels and garments. They worshipped the Lord with

heavenly parasols, banners, bells, flags, and with rows of lamps all around, and with manifold

kinds of worship. They played on heavenly musical instruments.‖ 108 The scripture also directs

that the material sūtra be elevated to demonstrate its transcendental status, as one might raise an

106
For more on icons, see Robert Sharf, ―Introduction: Prolegomenon to the Study of Japanese Buddhist Icon,‖ in
Living Images: Japanese Buddhist Icons in Context, ed. Robert H. Sharf and Elizabeth Horton Sharf (Stanford:
Stanford University Press, 2001), 1-18. And as Dobbins notes, the distinction between icon and relic is not always
apparent. See James C. Dobbins, ―Portraits of Shinran in Medieval Pure Land Buddhism,‖ in Living Images:
Japanese Buddhist Icons in Context, ed. Robert H. Sharf and Elizabeth Horton Sharf (Stanford: Stanford University
Press, 2001), 36-38. Bernard Faure and Robert Sharf have both published on the subject of icon, relic, portraiture,
and mummification. See Bernard Faure, The Rhetoric of Immediacy: A Critical Critique of Chan/Zen Buddhism
(Princeton: Princeton University Press, 1991) and Robert H. Sharf, "The Idolization of Enlightenment: On the
Mummification of Ch'an Masters in Medieval China," History of Religions 32, no. 1 (1992): 1-31. Sharf explores the
concept of ―icon of flesh‖, a translation from the Chinese meaning ―true-body portrait or flesh-body portrait‖ in this
article. He discusses the contradictions found in Buddhism of venerating bodily remains in the form of corporeal
relics and condemning the body as obscene and vulgar. He resolves this issue by providing evidence (and
extensively citing the scholarship of Gregory Schopen) that suggests relics, purified and transformed by the funeral
pyre, are no longer defiled, a concept that extends back to Vedic belief. Sharf also examines the contradiction
inherent in disciples—or the monk himself—trying to preserve the body after death. However, Sharf warns against
taking the canons as source material for daily life of the clergy and lay society. He discusses the precedents in
Chinese culture that could lead to Chan mummification, such as Han dynasty belief systems and Daoism. Sharf
examines the veneration of Chan mummies and the role of Chan portraiture in the commemorative process. See
Sharf, "On the Allure of Buddhist Relics," 75-99. In his book, Daitokuji, Gregory Levine argues that scholars have
neglected the role of visual mimesis in portraiture and ignored the culture of viewing. See Gregory Levine,
Daitokuji: The Visual Cultures of a Zen Monastery (Seattle: University of Washington Press, 2005), 71.
107
Conze, The Perfection of Wisdom in Eight Thousand Lines and Its Verse Summary, 289.
108
Ibid., 132.

152
icon upon a pedestal. Not only should one honor with luxurious materials and heavenly music

and establish the sūtra upon a lofty pedestal, the Diamond Sūtra proclaims that ―the spot of earth

where this Sūtra will be revealed, that spot of earth will be worthy of worship by the whole

world with its Gods, men and Asuras, worthy of being saluted respectfully, worthy of being

honored by circumambulation,—like a shrine will be that spot of earth.‖109 Eckel has pointed out

that, ―It is this causal association between the Perfection of Wisdom and the Buddha‘s

omniscience that makes it possible for the physical text to serve as ritual substitute for the

Buddha and to gather around itself all of the devotional actions normally associated with the cult

of the Buddha‘s relics.‖110

Several passages of the Lotus Sūtra reveal veneration of a Buddha through offerings of

great material and sensory value, which parallel the scripture‘s injunctions to worship the sūtra.

As an offering to the Buddha Pure and Bright Excellence of Sun and Moon, Medicine King

Bodhisattva (薬王菩薩 Jpn. Yakuō bosatsu, Ch. Yaowang pusa; Skt. Bhaiṣajyarāja bodhisattva),

in a former incarnation, ―entered into this samādhi, and in open space there rained down

māndārava and mahāmāndārava flowers, while a finely powdered, hard, black candana, filling all

the space, descended like a cloud. There also rained down the scent of candana of the near

seashore, the six shu [―weight equal to the twenty-fourth part of a tael‖]111 of this scent having

the value of the Sahā world [the secular world; 娑婆世界 Jpn. shaba sekai, Ch. suopo shijie]

sphere.‖112 Multiple passages convey similarly elaborate instructions for the appropriate

veneration of the sūtra after its proper transcription: ―If, having written down this scriptural roll,

he makes offerings with floral scent, necklaces, burned incense, powdered incense, perfumed
109
Conze, Buddhist Wisdom, 55.
110
Eckel, To See the Buddha, 97-98.
111
Digital Dictionary of Buddhism, accessed May 10, 2010, http://buddhism-dict.net.www2.lib.ku.edu:2048/cgi-
bin/xpr-ddb.pl?q=銖.
112
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 270. T. no. 262, 9: 53a29-b3.

153
paint, banners and parasols, garments, and sundry torches…the merit he gains shall also be

incalculable.‖113 The Lotus Sūtra again directs worshipers to treat the scriptural text as one

would the Buddha, revealing that those ―who shall look with veneration on a roll of this scripture

as if it were the Buddha himself, or who shall make to it sundry offerings of flower perfume,

necklaces, powdered incense, perfumed paste, burned incense, silk canopies and banners,

garments, or music, or who shall even join palms in reverent worship of it‖ 114 are, in point of fact,

honoring and worshiping the Buddha. These elaborate gifts befitting a Buddha as detailed here

correspond to the dedicatory rites of copied sūtras (Ten Kinds of Offerings, 十種供養 Jpn. jisshu

kuyō, Ch. shizhong gongyang) in which recently transcribed scriptures are presented with the ten

offerings. 115

Again, such praises and offerings worthy of the Buddha are also accorded to the Golden

Light Sūtra. Throughout the sixth chapter on the Four Guardian Kings, exalted perfumes,

heavenly odors, divine golden light, and brilliant umbrellas and banners honor this king of the

sūtras. But it is not just a matter of doctrine. Various and diverse early medieval records such as

the ones explored earlier in the chapter testify to the belief that dharma relics are imbued with the

potent power of the Buddha. Moreover, evidence from religious practices verifies that sūtras

were venerated for their potency and as relics of the Buddha. In the remainder of the chapter, in

order to demonstrate the practical application of the doctrinal assertion of sūtra text as dharma

relic, I examine the protocols for sūtra transcription and the messages communicated by the

materiality and visual description of the scriptures. I also explore the practice of sūtra burials and

113
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 274-75. T. no. 262, 9: 54b21-26.
114
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 159. T. no. 262, 9: 30c11-13.
115
Seki Hideo, Heian jidai no maikyō to shakyō 平安時代の埋経と写経, 4th ed. (Tokyo: Tōkyōdō Shuppan, 1999),
264. The ten offerings as recorded in the ―Preachers of Dharma‖ chapter of the Lotus Sūtra are flowers, ornaments,
incense, powdered incense, perfumed paste, burned incense, silk banners and canopies, clothing, music, and joined
hands in prayer.

154
the construction of dharma relic stūpas, as they reveal the treatment of scripture as dharma relics.

The combinatory practice merging sūtra and stūpa is also revealed to be a long tradition

informing the construction of the jeweled-stūpa mandalas.

Transcription of Scriptures: Revealing Sūtra as Dharma Relic

Because the practice of sūtra transcription has already been introduced in the second chapter, I

focus in this section on the procedural and visual aspects of scripture copies. The strict rules and

formality associated not only with copying, hearing, and expounding the scriptures, but also just

approaching the texts suggests the power of sacred word and the decorum required for respectful

engagement with the scriptures. Purity concerns dictate many of the requirements, such as the

stipulations in the Golden Light Sūtra instructing devotees in the proper way to approach the

sūtra: ―Having put on clean robes, wearing well-perfumed clothes, having produced a mind (full)

of love, one must do honour untiringly.‖ 116

The famous Tendai priest, Ennin 円仁 (794-864),117 is traditionally credited in early

medieval records, such as the thirteenth-century Important Documents of Mt. Hiei (叡岳要記

Eigaku yōki), with establishing the practice of copying sūtras in accordance with ritual

prescriptions (nyohōgyōhō Jpn. 如法経法, Ch. rufajingfa). In 833, feeling his body beginning to

fail and his eyesight diminish, Ennin retired to a grass hut in Yokogawa on Mt. Hiei to await

death; however, death did not come, and for three years he practiced austerities and meditated

116
Emmerick, The Sūtra of Golden Light, 2. T. no. 665, 16: 404b19-404b21.
117
For records of Ennin‘s travels in China, see Bussho Kankōkai 佛書刊行會, ed., ―Nittō guhō junrei gōki 入唐求
法巡礼行記,‖ in Dainihon bukkyō zensho 大日本佛教全書, vol. 113 (Tokyo: Bussho Kankōkai, 1912-1922), 169-
282. For an English translation, see Edwin O. Reischauer, Ennin's Diary: The Record of a Pilgrimage to China in
Search of the Law (New York: Ronald Press, 1955). For a brief biography of Ennin, Jikaku daishi den 慈覚大師伝,
written c. 970, see Saitō Enshin 斎藤円真, trans., Jikaku daishi den 慈覚大師伝 (Tokyo: Sankibō Busshorin, 1992).

155
while his health improved.118 For these three years (829-31), Ennin made preparations such as

growing his own hemp plants for the paper and ritualistically reciting the Lotus Sūtra each

morning, afternoon, and evening (the Lotus-confession rites commonly used in Tendai

Buddhism; 法華懺悔 Jpn. hokke zange, Ch. fahua chanhui) as absolution for his karmic debt

before finally copying the Lotus Sūtra in an exhausting ritual. 119 The purity of the materials is

paramount in the ritualistic transcription of sūtras, and Ennin went to great lengths to ensure that

the brush, ink, and paper were untainted by sins resulting from the use of animal products.

Instead of animal hairs, Ennin crafted a brush of grass and twigs; and rather than ink solidified

by animal-glue, he opted for graphite.120 Gifts of fragrant incense and flowers were made to each

character,121 much like the offerings made to an icon of the Buddha, such as the ten offerings. In

1031, these painstakingly transcribed scrolls were buried in Yokogawa inside a copper container

by the monk, Kakuchō 覚超 (960-1034).122

While Ennin‘s preparation and reproduction of the scripture (nyohōgyō 如法経) laid the

precedent for ritualistic transcription in the Tendai school, the extensive procedures were

obviously abbreviated. Tanaka Kaidō examines early medieval texts to understand the rigourous

ritualistic transcription.123 The first seventeen days are spent cleansing the body and spirit by

penitence and fasting.124 During the next twenty-seven days, the paper and water to be used in

118
Mochizuki Shinkō 望月信亨, Mochizuki bukkyō daijiten, vol. 5 (Tokyo: Sekai Seiten Kankō Kyōkai, 1966-68),
4140.
119
Kageyama Haruki 景山春樹, ―Yokogawa ni okeru nyohō shakyō to maikyō 横川における如法写経と埋経,‖
Kōkogaku zasshi 考古学雑誌 54 (1969): 3.
120
Mochizuki, Mochizuki bukkyō daijiten, 4140.
121
Seki, Heian jidai no maikyō to shakyō, 264.
122
Ibid., 264.
123
Tanaka, Shakyō nyūmon, 40-46.
124
Tanaka notes that this initial step is crucial for first time copyists, but may be omitted for those who routinely
practice ritualistic transcription. See Ibid., 40.

156
the copying are carefully prepared, involving rituals around the altar space. 125 The altar space

should be specifically crafted: the four corners of the altar should have flower vases and

offerings such as burning incense and a canopy overhead. The table upon which the sūtra is

copied should also have incense with all sorts of decorative accents like banners and nets made

from strings of jewels. 126 After announcing to the main icon that the ceremony is about to begin,

the Lotus Sūtra is recited aloud, sins are repented, and full body prostrations are made. 127 The

prescription for the preparation of the ink strictly dictates that, having applied incense to the

body, the ink should be ground and filtered through a cloth. The drops are collected into a bowl

and this process is repeated until enough pure ink has been gathered. The papers used for the

transcription are then glued together using the root of grass as an adhesive. 128 All these steps are

merely the preparatory measures for ritualistically transcribed sūtras. The procedure of copying

often prominently involves the body of the copyist: after writing the first character or the first

line of characters, five bodily prostrations are made followed by three bows to the characters, a

method known as ichiji sanrei.129 At the close of the transcription project, the sūtras can be

enshrined within containers for burying (described below) and the ten offerings made to it. 130

The meticulous preparation and transcription of the sūtras reflect the great reverence for sacred

word.

Laborious and elaborate copying rituals were pursued in China as well. For example,

Stevenson notes that the monastic historian Zanning 贊寧 (919-1001) observes of contemporary

transcription practices,

125
Ibid., 42.
126
Ibid., 40-41.
127
Ibid., 41-42.
128
Ibid., 42.
129
Ibid., 43. For more on ichiji sanrei, see chapter two of this dissertation.
130
Ibid., 44.

157
 ‗There are persons who, in imitation of the ancients, venerate the texts of the Lotus and
Flower Garland sūtras [by prostrating themselves to] each character, one at a time. They
regard this to be veneration of the undefiled treasure-store of the dharma itself. Thus we
find members of the fourfold saṅgha who actually insert the words ―homage to‖ [namo]
before each word and ―-buddha‖ [fo] after each word [of the sūtra].‘131

Daniel Veidlinger, quoting a late fourteenth-century Southeast Asian chronicle written by

Dhammakitti, Saddhammasangaha, provides another parallel, albeit later, example of venerating

sūtra text as the Buddha: ―each letter should be considered as a Buddha image, therefore the wise

should write the Tipiṭaka.‖132

The practice of burying sūtras in preparation for the return of the future Buddha,

Maitreya (弥勒菩薩 Jpn. Miroku bosatsu, Ch. Mile pusa), at a time when the dharma will have

all but vanished, corresponding to the last phase of our world-age known as mappō, represents

another widespread religious practice venerating, and more importantly, preserving sūtras as

dharma relics.133 Characterized by D. Max Moerman as ‗the archaeology of anxiety,‘ kyōzuka 経

塚 (burial of sūtras in reliquary mounds) were intended to preserve the Buddha‘s dharma, which

would well up out of the ground at the advent of Maitreya; but they also were interred with

fervent prayers for personal salvation, the birth of heirs, and even cures for minor physical

afflictions.134 Significant to this study is the method in which the buried sūtras (埋経 Jpn. maikyō,

Ch. maijing) were interred. The scriptures, copied on paper often in the ritualistic manner of

transcription (nyohōgyōhō), but also on more permanent materials such as tiles, wood, copper,

131
Stevenson, ―Buddhist Practice and the Lotus Sūtra in China,‖140. This quote comes from the Song Version of the
Biographies of Eminent Monks (宋高僧傳 Jpn. Sō kōsō den, Ch. Song gaoseng zhuan) by Zanning. T. no. 2061,
50:888b13-16.
132
Daniel M. Veidlinger, Spreading the Dhamma: Writing, Orality, and Textual Transmission in Buddhist Northern
Thailand (Honolulu: University of Hawai‘i Press, 2007), 177.
133
Sekine Daisen 関根大仙, Mainōkyō no kenkyū 埋納経の研究 (Tokyo: Ryūbunkan, 1968), 108-81. For a very
good introduction to the practice of kyōzuka with many images, see Seki, ―Kyōzuka to sonno ibutsu.‖
134
D. Max Moerman, ―The Archeology of Anxiety: An Underground History of Heian Religion,‖ in Heian Japan:
Centers and Peripheries, ed. Mikael Adolphson, Edwards Kamens, and Stacie Matsumoto (Honolulu: University of
Hawai‘i, 2007), 245-271.

158
stone slabs, and seashells, 135 were sealed in sūtra cylinders (経筒 Jpn. kyōzutsu, Ch. jingtong).

The small reliquaries entombing the dharma relics varied in form from the most simple and

understated to elaborate jeweled stūpas. Buried in small, stoned-lined underground chambers

often filled with charcoal to combat water damage, sealed with stone and marked by a raised

earthen mound and, at times, a stone stūpa, the entire process reflects not only the anxiety over

preserving the dharma (and one‘s own salvation), but a profound reverence for sūtra text as relics

of the Buddha. 136

And as J. Edward Kidder points out, ―it was not unusual to include a sword or two in the

mound as protection for holy texts.‖137 Often times the sūtra containers themselves were

inscribed with protective phrases such as calling on the name of the Lotus Sūtra (namu

myōhōrengekyō) and the mantra of light (光明真言 kōmyō shingon).138 On a sūtra container

commissioned by Fujiwara Michinaga discovered on Mt. Kinpu 金峰山 and dated 1007, the

inscription declares: ―Burying the relics of the dharmakāya recalls Śākyamuni‘s mercy.‖ 139

Clearly, the handling of sūtra text in this archaeological context reflects the many injunctions in

various scriptures to not only copy and worship the text, but also to enshrine it as dharma relics

and as embodiments of the dharmakāya principle.

These representative examples offer a glimpse into the elaborate preparatory methods of

sūtra transcription, whose complex care and concern reflect the nonduality of sacred texts and

the Buddha. Visually, this fundamental conflation is manifested again and again in extant sūtra

135
Yamakawa Kumiko 山川公見子, ―Kyōzuka no shinkō 経塚の信仰,‖ Kikan kōkogaku 季刊考古学 97 (2006):
72.
136
For an interesting, but later example of a kyōzuka with relics in the form of precious gems and colored glass as
well as a sūtra, see Yajima Kyōsuke 矢島恭介, ―Konshi kinjikyō to busshari 紺紙金字経と仏舎利,‖ Museum 81
(1957): 2-6.
137
J. Edward Kidder, Jr., ―Busshari and Fukuzō: Buddhist Relics and Hidden Repositories of Hōryūji,‖ Japanese
Journal of Religious Studies 19 (1992): 224.
138
Rambelli, Buddhist Materiality, 109.
139
Seki, ―Kyōzuka to sonno ibutsu,‖ 19. For an image, see Ibid., 1 fig. 1.

159
copies. In the eleventh-century Ichiji butsu hokekyō at Zentsūji, the copyist has graphically

illustrated this concept by placing a figure of the Buddha beside each character, thereby

rendering visual the nonduality of the Buddha and his word. One cannot see or read the Ichiji

butsu hokekyō without registering the conflation of sūtra text and Buddha reinforced by the

symbolically emblematic placement of a Buddha figure next to the letters of the sacred scripture.

Scrolls such as these were copied with faith that the copyist/patron will be reborn in paradise and

also that worldly benefits, such as good health, protection, and material rewards, will be

granted.140 Similar treatment of sacred word is seen in scrolls where each character is supported

by a lotus pedestal, such as the Ichiji rendai hokekyō of the eleventh or twelfth century. The

enshrinement of individual letters upon a pedestal recalls the ubiquitous practice, visualized in

countless sculptures and paintings, of elevating a Buddha on the pure seat of a lotus. This

deferential treatment of enlightened beings stems from doctrinal descriptions of preaching

Buddhas. Thus not only do such scrolls visually conjure the nonduality of dharma and Buddha

by borrowing from established iconography, the mode of representation implies a deep reverence

for dharma relics by establishing the sacristy of each character. Furthermore, the twelfth-century

Lotus Sūtra handscrolls adorn each textual character with a stūpa as an example of the Ichiji hōtō

hokekyō format. Enclosing the word within a reliquary enshrines—in a very literal way—the

relics of the Buddha. These particular scrolls also represent a unification of dharma relic and

stūpa, albeit somewhat less involved than the relationship imagined in the jeweled-stūpa

mandalas. The mandalas represent one of the most complex and multifaceted visual treatises on

the nature of dharma, relic, body, and sūtra. Whereas the handscrolls lend visual description to

the notion of dharma as relic, the mandalas not only privilege the text of the sūtras as the

centerpiece of the paintings and serve as visual commentaries on the nature of sūtra as dharma
140
Tanaka, Nihon shakyō sokan, 20-22.

160
relic by referencing the conventional partnering of relic and reliquary; they further layer the

symbolism of scriptural text as the body of the Buddha through the structuring of the dharma in

the form of a stūpa.

At the foundation of this visual culture and religious practice is what Schopen has

described as the ‗cult of the book.‘141 Schopen reveals that early Mahayana groups developed

around particular sūtras, promoting a ‗cult of the book‘ often in contrast to and stressed above

the focus on relics and stūpas. As he points out, ―this cult did not develop in isolation, it had to

contend at every step with the historical priority and the dominance of the stūpa/relic cult of

early Buddhism in the milieu in which it was attempting to establish itself.‖ 142 Schopen‘s study

confirms the doctrinal foundation in early texts for classifying textual dharma as relics of the

Buddha—dharma relics which, in places the book was venerated, established a sacred space,

variously characterized as a caitya (支提 Jpn. shidai, Ch. zhiti) or stūpa. Thus the cult of the

book borrows the language of sacristy and metaphors of authority and salvific power from the

cult of the stūpa/relic in establishing its superiority. And invariably, sūtras that privilege the cult

of the book negatively assess corporeal relics and related stūpa construction. The third chapter of

the Perfection of Wisdom in Eight Thousand Lines offers a prototypical example of the

privileging of the cult of the book:

Suppose that there are two persons. One of the two, a son or daughter of good family, has
written down this perfection of wisdom, made a copy of it; he would then put it up, and
would honour, revere, worship, and adore it with heavenly flowers….The other would
deposit in stūpas the relics of the Tathagata who has gone to parinirvana; he would take
hold of them and preserve them; he would honor, worship and adore them with heavenly
flowers….Which one of the two, O Lord, would beget the greater merit?....The son or
daughter of good family who has made a copy of the perfection of wisdom, and who

141
Gregory Schopen, ―The Phrase ‗Sa Pŗthivīpradeśaś Caityabhūto Bhavet‘ in the Vajracchedikā: Notes on the Cult
of the Book in Mahāyāna‖ Indo-Iranian Journal 17 (1975): 147-81.
142
Schopen, ―The Phrase ‗Sa Pŗthivīpradeśaś Caityabhūto Bhavet‘ in the Vajracchedikā,‖ 168.

161
 worships it, would beget the greater merit. For by worshiping the perfection of wisdom
he worships the cognition of the all-knowing.143

Combinatory Practices of Sūtra and Stūpa

The transcription of sacred text was an ubiquitous practice. In this section, I demonstrate that it

was also an amalgamated one, in which the copying of sūtra was often not the sole pursuit.

Sūtras were frequently paired with stūpas in a variety of ways, manifesting the understanding of

scripture as dharma relics and revealing the polyvalent notions of the bodies of the Buddha as

revealed through both sūtra and stūpa, a topic continued in the next chapter. The jeweled-stūpa

mandalas embody an innovative format of sūtra transcription: the central icon as an inventive

structure carrying meaning and marking an iteration in the long history of the combination of

sūtra and stūpa in visual culture, religious practice, and doctrine.

The religious practice pairing sūtra text and stūpas was established long ago. Both Faxian

(法顯 337–c.422) and Xuanzang (玄奘 602-664) bear witness in their travel diaries to the

practice of dharma relic stūpas. In the text, Record of Buddhist Countries (佛國記 Jpn. Bukkoku

ki, Ch. Foguoji), Faxian records in his visit to India during 399-414 that stūpas were constructed

for specific purpose of sūtra veneration: ―Where a community of monks resides, they erect topes

[stūpas] to … the sūtras [經塔 Jpn. kyōtō, Ch. jingta].‖144 Faxian‘s testimony perhaps introduced

the practice to China. Xuanzang likewise records the ubiquitous and related practice of

enshrining sūtra verses in mini-stūpas:

It is a custom in India to make little stūpas [小窣堵波 Jpn. shō sotoba, Ch. xiao sudubo;
Skt. ] of powdered scent made into a paste; their height is about six or seven inches, and

143
Conze, The Perfection of Wisdom in Eight Thousand Lines and Its Verse Summary, 105-06.
144
James Legge, A Record of Buddhist Kingdoms: Being an Account by the Chinese Monk Fa-hien of His Travels in
India and Ceylon in Search of the Buddhist Books of Discipline (New York: Dover, 1965[1886]), 44-45. T. no. 2085,
51: 859b18-9.

162
 they place inside them some written extract from a sūtra; this they call a dharma śarīra
(fa-shi-li [sic]) [dharma relic; 法舎利 Jpn. hōshari, Ch. fasheli]. When the number of
these has become large, they then build a great stūpa, and collect all the others within it,
and continually offer to it religious offerings. This then was the occupation of Jayasēna
(Ching-kian); with his mouth he declared the excellent law, and led and encouraged his
students, whilst with his hands he constructed these stūpas. Thus he acquired the highest
and most excellent religious merit.145

As archaeological evidence has produced in abundance, the ye dharmā hetuprabhavā verse

inscribed on clay seals was commonly enshrined in such stūpas. And as Boucher points out, the

Sūtra on the Merit of Building a Stūpa as Spoken by the Buddha encourages the construction of

these dharma śarīra stūpas (dharma relic stūpas; 法舎利塔 Jpn. hōshari tō, Ch. fasheli ta),

expounding on the great merit accrued from such devotional acts. The sūtra employs the standard

formula of question and answer session: Avalokiteśvara asks the Buddha the proper method for

stūpa construction to which the Buddha responds that if one were to build a stūpa, regardless of

its size, on a previously unestablished place ―and if inside this stūpa one encloses the [body of

the] Tathāgata down to even one minute portion of his relics, hair, teeth, beard, or fingernails; or

else if one deposits the twelve section scripture, which is the storehouse of the Tathāgata‘s

dharma, down to even one four line verse, this person‘s merit will be as great as the brahma

heaven.‖146 The Buddha then clarifies that the ‗one four line verse‘ is the ye dharmā

hetuprabhavā verse, of which he reveals, ―this verse signifies the Buddha-dharmakāya. You

should write [this verse] and place it inside the stūpa. Why? Because all causes and the dharma-

nature of all things that are produced are empty. This is the reason that I call it the dharmakāya.

If a living being understood the import of such causes, you should know that this person would

then see the Buddha.‖147

145
Samuel Beal, trans., Si-yu-ki: Translated from the Chinese of Hiuen Tsiang (A. D. 629), vol. 2 (London: Trübner
& Co., 1884), 146-47. T. no. 2087, 51: 920a21-6.
146
Boucher, ―Sūtra on the Merit of Bathing the Buddha,‖ 9. T. no. 699, 16: 801a28-b2.
147
Boucher, ―Sūtra on the Merit of Bathing the Buddha,‖ 9-10. T. no. 699, 16: 801b12-4.

163
 This practice, which seems to have been widespread in India, carried over to China where

it met with enthusiasm. Hsueh-man Shen has analyzed Chinese relic deposits, including dharma

relic stūpas, from the seventh century to the mid-twelfth century. 148 She also finds evidence of

great numbers of stūpas dedicated to the enshrinement of textual dharma as not only the

dharmakāya of the Buddha, but as expressions of the entire body of the Buddha. On a protective

cloth ensconcing a copy of the Lotus Sūtra discovered within a Liao deposit is written, ―the

whole of the Lotus Sūtra as the entire-body śarīra is in this pagoda.‖149 Furthermore, she finds

evidence at Shijiafoshelita 释迦佛舍利塔 that more than just the entire body of the Buddha, the

scriptures embody the three bodies—thus the entire body—of all Buddhas, past, present, and

future. The Fo xingxiang zhong anzhi fasheli ji or Instructions in Enshrining Dharma-śarīra

inside Buddha Images, likely a Liao compilation excavated at the site, quotes the Dhāraṇī of the

Seal on the Casket [of the Secret Whole-body Relic of the Essence of All Tathāgatas]:150 ―The

Buddha told Vajrasattva Bodhisattva that the entire-body śarīras of all Buddhas, including those

of the future, of the past, and those who entered nirvāna, all exist in the Baoqie yin tuoluoni

dhāraṇī [an abbreviation of the longer sūtra title]. All these Buddha‘s three bodies are also

present in it.‖151 While such distillations were commonly used in esoteric rituals, in cases like

these, they were commissioned to establish sites of sacristy. Shen demonstrates through the

analysis of several more texts, inscriptions, and other findings at relic deposits that what began in

India remained a popular devotional practice in China. In this we see a three-dimensional parallel

to the jeweled-stūpa mandalas: the dharma relic stūpas (or as they are occasionally called, fashen

148
Hsueh-man Shen, ―Realizing the Buddha‘s Dharma Body During the Mofa Period: A Study of Liao Buddhist
Relic Deposits,‖ Artibus Asiae 61, no. 2 (2001): 263-303.
149
Ibid., 271. This echoes a passage from the ―Apparition of the Jeweled Stūpa‖ chapter in the Lotus Sūtra.
150
Jpn. Issai nyorai shin himitsu zenshin shari hōkyōin darani kyō; Ch. Yiqie Rulai xin mimi quanshen sheli baoqie
yin tuoluoni jing; Skt. Sarvatathāgatādhiṣṭhā hṛdaya guhya dhātu karaṇḍa mudrā nāma dhāraṇī; 一切如來心祕密
全身舍利寶篋印陀羅尼経; T. no. 1022a, 19: 710a10-712b7.
151
Shen, ―Realizing the Buddha‘s Dharma Body During the Mofa Period,‖ 272.

164
shelita 法身舎利塔, which places emphasis on ‗body,‘) create through the enshrinement of

sūtras or verses an architectural dharmakāya; in a related manner, the mandalas manifest through

the imbrication of scriptural text and stūpa the different, yet ultimately conflated, notions of body

through architextual expression. The enclosing of dharma relics in stūpas also found expression

in Korea where the practice included some interesting combinatory aspects, such as the

installation of both textual and corporeal relics in a stūpa deposit;152 the same practice is also

found in early medieval Japan. Constructing architectural stūpas for the enshrinement of sūtras

has a long history in Japan. For example, Kawakatsu Kenryō, in his study on the origins and

many manifestations of the many jeweled-stūpas (多宝塔 tahōtō) and its connections to the

Lotus Sūtra, notes that during the ninth century Saichō, in a not uncommon display of veneration,

commissioned several stūpas that were each to enshrine a tremendous thousand copies of the

Lotus Sūtra.153

During the Goryeŏ dynasty (918-1392) mini-stūpas housing dhāraṇ īs were

commissioned,154 a practice consistent with Chinese precedence and one that took root in Japan

as well. From 764-70, Empress Shōtoku 称徳天皇 (718-70) commissioned the remarkable

project of one million mini-stūpas (百萬塔 Jpn. hyakumantō)155 containing an assortment of

dhāraṇ ī156 from the scripture, Dhāraṇī of the Pure Immaculate Light,157 to be donated to several

152
Jan Fontein, ―Śarīra Reliquary from Pagoda of Powŏn-sa Temple Site,‖ Misul charyo 47 (1991): 104-08. For
more on Korean reliquaries, see National Museum of Korea, ed., Pulsari changom (Seoul: National Museum of
Korea, 1991). For more information on Korean relic deposits within stūpas, and the variations therein, and within
sculptural icons, see Umehara Sueji 梅原末治, ―Kankoku keishū kōfukuji tō hakken shari yōki 韓国慶州皇福寺塔
発見の舎利容器,‖ Bijutsu kenkyū 美術研究 156 (1950): 31-47.
153
Kawakatsu Kenryō 川勝賢亮, ed., Tahōtō to Hokekyō shisō 多宝塔と法華経思想, (Tokyo: Tōkyōdō, 1984): 52-
55. For a brief examination into the origins of faith in Tahō, see Ocho Enichi 横超慧日, ―Tahōtō shisō no kigen 多
宝塔思想の起源,‖ Indogaku bukkyōgaku kenkyū 印度学仏教学研究 2, no. 1 (1953): 30-36.
154
Hye-Bong Ch‘ŏn, ―Dhāraṇī-Sutra of the Early Koryŏ,‖ Korea Journal 12 (1972): 4-12.
155
Ishida Mosaku, ―Tō, tōba, sutsūpa 塔 塔婆・スツーパ,‖ Nihon no bijutsu 日本の美術 77 (1972): 50 fig. 98.
156
While the sūtra provides six possible dhāraṇī, only the following have been discovered: the Konpon darani 根本
陀羅尼, Sōrin darani 相輪陀羅尼, Jishin’in darani 自心印陀羅尼, and Rokudo darani 六度陀羅尼.

165
influential temples. 158 Throughout the short sūtra, grand promises are made: for example, the

prolongation of life, release from horrible rebirths, such as in the realm of the hungry ghosts, the

absolution of all sins, eradication of evil, and crucial to an imperial commission, the protection of

the sovereign and the nation if projects like the one million mini-stūpas (and even far less

grandiose projects) are sponsored using the sūtra‘s dhāraṇ ī. A related practice known as momitō

籾塔 (rice-grain stūpas) wraps an unhulled grain of rice inside a small slip of paper upon which

lines of the Dhāraṇī of the Pure Immaculate Light are written and enshrines the rolled paper

within small votive stūpas, usually still bearing the marks of the knife that carved them. A vast

store of fifteenth-century momitō were discovered under the altar of the mirokudō 弥勒堂 of

Murōji 室生寺.159 Inside fifteen or sixteen straw sacks, over 37,000 momitō of plain and colored

wood were unearthed.160 Sherry Fowler explains that the construction of momitō was a regional

trend revealing the area‘s focus on agricultural productivity. The commission of the stūpas is

strongly linked to wishes for good harvests with their connection to rain and the visual

resemblance of the grain of rice to relics and the consequent association of the worship of

jewels/relics with fecundity, including that of the harvests.161

The combinatory practices bringing together sūtra and stūpa also took inventive forms

during the eleventh through thirteenth centuries. Rather than the more common combination

enshrining sūtra within stūpa like the examples provided above, a classification of objects known

as tōkyō (‗stūpa-sūtras;‘ 塔経 Ch. tajing) reflect a step further along the imbrication scale. The

157
Jpn. Mukujōkō daidarani kyō; Ch. Wugoujingguang datuoluoni jing; Skt. Raśmivimalaviśuddhaprabhā-dhāraṇī;
無垢淨光大陀羅尼経; T. no. 1024, 19: 717c7-721b22.
158
For a discussion of the hyakumantō, see Brian Hickman, ―A Note on the Hyakumantō Dhāranī,‖ Monumenta
Nipponica 30, no. 1 (1975): 87-93; and Mimi Hall Yiengpruksawan, ―One Millionth of a Buddha: The Hyakumantō
Darani in the Scheide Library,‖ The Princeton University Library Chronicle 48, no. 3 (1987): 224-38.
159
Sherry D. Fowler, Murōji Rearranging Art and History at a Japanese Buddhist Temple (Honolulu: University of
Hawai‘i Press, 2005), 27.
160
Ibid., 29 fig 1.9.
161
Ibid., 30-33, and 112-14 for further examinations of the momitō.

166
category of tōkyō is large and ambiguous; the jeweled-stūpa mandalas have even been called

tōkyō. Two of the more pervasive examples of the tōkyō are deitōkyō 泥塔経 (small clay stūpas

onto which a character from a sūtra is inscribed; Ch. nitajing) and kokerakyō 杮経 (strips of

wood in the shape of stūpas with inscriptions of sūtra text), although another well-known type is

the Ichiji hōtō hokekyō handscrolls so classified because the scrolls enshrine the text inside

stūpas. The Lotus Sūtra is at the center of the scriptures manipulated in the tōkyō formats.162

Deitōkyō, molded from simple materials such as pastes made of incense, clay, sand, and

mud, have a long history of production in India, China, Korea, and Japan. Xuanzang‘s record

testifies to the practice in India. Sometimes passages from scriptures were rolled and placed

inside the deitōkyō, but the examples examined in this section usually offer single textual

characters upon the body of the stūpa.163 Huge numbers of deitōkyō were produced per

commission, and the anticipated benefits from these extensive projects ranged from prolonged

life, cures for illness, and even spirit appeasement.164 Brian Ruppert, noting an increase in the

number of deitōkyō commissioned during the early 1240s, hypothesizes that the clay stūpas were

attempts to pacify the spirit of the cloistered Emperor Go-Toba who died while in exile. 165

Kokerakyō have been known by various names causing some confusion. They have been

variously referred to as sasatōba 笹塔婆 (a memorial stūpa made of bamboo grass), mokkan

shakyō 木簡写経 (sūtra copying upon long, wooden strips), and senbon tōba 千本塔婆 (one

thousand stūpas), a reference to the large quantity of each kokerakyō project.166 The term sotōba

卒塔婆 (stūpa) can be used to subsume the entire category of stūpa construction from the slender

162
Miya, Kinji hōtō mandara, 2.
163
For images, see Ishida, ―Tō, tōba, sutsūpa,‖ 58-59 figs. 127-134.
164
Ibid., 58.
165
Ruppert, Jewel in the Ashes, 239.
166
Ishida, ―Gangōji gokurakubō hakken no kokerakyō,‖ 230.

167
kokerakyō to full-sized stūpas of temple architecture.167 The practice of kokerakyō was the most

popular during the Heian and Kamakura periods with examples remaining from the Muromachi

to Edo periods, but they are much rarer.168 The sizes of kokerakyō are roughly 25-30 cm long, 1-

1.5 cm wide, and 1.5 cm thick.169 Ishida Mosaku identifies three common shapes of kokerakyō:

the top of the strip is cut into a triangular shape; the top is cut into a triangular shape with

indented sides to more resemble a stūpa; and the top is shaped like a gorintō 五輪塔 (a five-

ringed stūpa).170 The earliest mention of kokerakyō in early medieval documents comes from the

Hyakurenshō, a thirteenth-century anthology of various records and tales by an unknown

compiler. In the tenth month and eleventh day of 1181, Hyakurenshō records that Taira

Shigemori 平重盛 (1138-79) told Go-Shirakawa of his dream in which one thousand volumes of

the Heart Sūtra were copied onto kokerakyō in order to pacify the troubled spirits of the

Heike. 171 Learning this dream, Go-Shirakawa then commissioned the project and accumulated

twelve barrels of kokerakyō, setting them adrift upon the east and west seas. Tanaka Kaidō notes

that rather than interpret kokerakyō as persimmon leaves because of the vagueness of the passage

and because persimmon and kokera share the same kanji (杮), he points out that by the tenth

month, the leaves of the persimmon tree have already fallen and gone, thus making the use of

persimmon leaves highly unlikely. 172 Ōta Masahiro 太田正弘 notes that if the kokerakyō

contained the standard seventeen characters, then it would take nineteen such lines to copy the

short Heart Sūtra, thereby resulting in 19,000 pieces for the one thousand copies of the sūtra. 173

167
Ōta Masahiro 太田正弘, ―Kokerakyō ni tsuite: Aichiken Iwakurashi Taisanji shutsudohin o chūshin to shite 杮経
について:愛知県岩倉市大山寺出土品を中心として,‖ Kokugakuin zasshi 国学院雑誌 76 (1975): 32.
168
Ōta, ―Kokerakyō ni tsuite,‖ 32.
169
For images of kokerakyō, see Ishida, ―Tō, tōba, sutsūpa,‖ 54 figs. 112-15, 56 figs. 119-20 and 122.
170
Ishida, ―Gangōji gokurakubō hakken no kokerakyō,‖ 230.
171
Tanaka, Nihon shakyō sokan, 28. Kuroita, ―Hyakurenshō,‖ 105.
172
Tanaka, Nihon shakyō sokan, 28.
173
Ōta, ―Kokerakyō ni tsuite,‖ 31.

168
 The earliest material example of kokerakyō is an example dated 1215 and discovered in

the attic of Gangōji‘s 元興寺 gokurakubo 極楽坊, the living quarters of the monks at Gangōji in

Nara.174 Ishida‘s study of the kokerakyō uncovered at the temple revealed the method of

production.175 Normally, kokerakyō contained sūtra text on the front and back, but the text on the

front and back of one piece does not usually show any connection. But, among the 20-30,000

pieces of kokerakyō discovered at Gangōji, five sets of twenty kokerakyō banded together with

rolled paper binding were uncovered. The researchers noticed that unlike the disbanded and

scattered kokerakyō typically found, they were able to read the text of the wooden strips

continuously, beginning with the front and continuing onto the back. From these joined sets, the

method of kokerakyō transcription was revealed. Twenty pieces of kokerakyō were laid out, text

copied upon the front side, and then the pieces were turned over and the backs copied (beginning

with the last front piece copied), after which they were banded together. Ishida believes that the

kokerakyō were dedicated in this banded manner.176 Surveying the content of the pieces, the

researchers discovered that the overwhelming majority of the pieces containing sūtra text were

copied from the Lotus Sūtra.177 This is only one example of a very popular transcription method.

Other instances can be seen at Risshakuji 立石寺 in Yamagata prefecture, Chūsonji in Iwate

prefecture, and Kamakura‘s Kakuonji 覚園寺.

An early textual source for the transcription of kokerakyō comes from the diary of

Fujiwara Tametaka 藤原為隆 (1070-1130), Eishōki 永昌記. In the fourth month and eleventh

day of 1107, Tametaka records that the Lotus Sūtra was written in one day on in the shape of a

174
Ishida, ―Gangōji gokurakubō hakken no kokerakyō,‖ 232.
175
Ibid., 231.
176
Ibid., 234.
177
Ibid., 236.

169
stūpa.178 But given the ambiguity of the entry, the record might be referring to the copying of the

Lotus Sūtra in the Ichiji hōtō hokekyō format, or another method of transcription. The crucial

part is that this early medieval diary speaks to the combinatory practice of sūtra and stūpa. Other

early medieval records also document the merging of sūtra and stūpa in religious practice. For

instance, the entry for the eight month of 1140 from the Journal of the Monk Sainen on the

Memorial Service with Blue-Gold Sūtras (僧西念紺紙金字供養日録 Sō Sainen konshi kinji

kuyō nichiroku) records the construction of 69,384 sotōba and another dedication of gold and

silver inscribed deitōkyō of multiple sūtras along with a blue and gold Lotus Sūtra, totaling

97,189 characters and stūpas.179 The ambiguity of the term sotōba makes it difficult to determine

the exact nature of some of the commission, beyond the combination of sūtra and stūpa. Many

early medieval texts simply mention the construction of sotōba with no further clarification. The

late Heian and early Kamakura text, Sanbutsu jōshō 讃仏乗抄, records the commission of

sotōba on the twenty-first day of the eighth month in 1187 as well the transcription of the Lotus

Sūtra onto sotōba, among several other projects, as a dedication to the patron‘s deceased wife on

the twenty-second day of the fifth month in 1193. 180

As evidenced above, the combinatory practice utilizing sūtra and stūpa was a common

and long-standing tradition across much of East Asia. As many scholars have discussed, the

desire to combine sūtras and stūpas in one project likely stemmed in great part from the merit

generated from the conflation of the two highly meritorious forms of devotion. Sūtras

178
Ibid., 230. See Fujiwara Tametaka 藤原為隆, ―Eishōki 永昌記,‖ in Yōmei sōsho 陽明叢書, ed. Yōmei Bunko 陽
明文庫, vol. 29 (Kyoto: Shibunkaku Shuppan, 1988), 84.
179
Ishida, ―Gangōji gokurakubō hakken no kokerakyō,‖ 230. Sainen 西念, ―Sō Sainen konshi kinji kuyō nichiroku
僧西念紺紙金字供養日録,‖ in Insei jidai no kuyō mokuroku 院政時代の供養目録, ed. Miyake Yonekichi 三宅米
吉 and Tsuda Noritake 津田敬武 (Tokyo: Tokyo Teishitsu Hakubutsukan, 1924), 17 and 32-33.
180
Miya, Kinji hōtō mandara, 2. Fujita Tsuneyo 藤田経世, ed., ―Sanbutsu jōshō 讃仏乗抄,‖ Kōkan bijutsu shiryō
校刊美術史料, vol. 3 (Tokyo: Chūō Kōron Bijutsu Shuppan, 1975), 87-88.

170
commanded to be copied and promised great rewards for doing so. Komatsu Shigemi reveals that

the Lotus Sūtra accounts for approximately ninety percent of all surviving scriptures from the

Heian period.181 This owes in part to the several instances within the sūtra that instruct devotees

to copy its text and disseminate the dharma, resulting in great rewards for the practitioner:

[I]f a good man or good woman shall receive and keep, read and recite, explain, or copy
in writing a single phrase of the Scripture of the Dharma Blossom, or otherwise and in a
variety of ways make offerings to the scriptural roll with flower perfume, necklaces,
powdered incense, perfumed paste, burned incense, silk banners and canopies, garments,
or music or join palms in reverent worship, that person is to be looked up to and exalted
by all the worlds, showered with offerings fit for a Thus Come One [a Buddha]. 182

Hence, the redemptive power of the Lotus Sūtra is so great that to copy or intone even one

phrase is to gain the status of the Buddha. And as will be examined in the next chapter, similar

injunctions are made for stūpa construction. Ishida Mosaku explains that four merit-generating

methods have characterized Buddhism: making banners, constructing stūpas, copying scriptures,

and carving sculptures.183 He posits that in early forms of Buddhism, banners played a crucial

role as the symbol of the religion and that stūpas, after the parinirvāṇa, became the symbol of

the Buddha and served to expand the religion along with banners. Copying scriptures was

important not only for the dissemination of the faith, but also as a meritorious activity, much like

sculptures. Ishida notes that from the Heian period on, attempts were made to combine some of

the four types of activities in one project: banners with the image of a Buddha, sūtras placed

within sculptures, sūtra copies of alternating lines of script and images of Buddhas (Ichiji butsu

hokekyō), and tōkyō such as kokerakyō.184 The merit is thus doubled and with only marginal

effort and expense expended compared to the commission of individual projects. Miya Tsugio

181
Komatsu, Heike nōkyō no kenkyū, vol. 1, 47.
182
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 160. T. no. 262, 9: 30c17-21.
183
Ishida, ―Gangōji gokurakubō hakken no kokerakyō,‖ 229.
184
Ibid.

171
claims that the jeweled-stūpa mandalas manifest the meritorious activities of building stūpas,

copying sūtras, and interpretation of the dharma. 185

While not isolated in the combination of text and reliquary, the jeweled-stūpa mandalas

represent an innovative leap in obeying the many commands across numerous texts to construct

stūpas and copy sūtras. Not only do the mandalas fulfill the injunction to honor, revere, and copy

the scriptures, thereby reaping considerable salvific benefit; the mandalas also metaphorically

manifest the injunction to erect stūpas. But where once reliquary contained relic, guarding and

hiding it from sight; through a conceptual twist, relic now constructs reliquary in the jeweled-

stūpa mandalas, thereby conflating the two as one. The mandalas perhaps embody a more

economical manifestation of the enjoiner to erect architectural reliquaries—not always a

financially feasible option. The Lotus Sūtra is celebrated for its unifying perspective on both the

cult of the stūpa and the cult of the book; and, as the Lotus Sūtra is the most commonly used

sūtra in the jeweled-stūpa mandalas format, it stands to reason that this rather equitable

confirmation of both devotional practices did not go unnoticed. At multiple points the sūtra

proclaims the transcendent value of both devotional activities, comparing the merit and rewards

generated from copying the sūtra and erecting stūpas to the Buddhas and suggesting a nondual

parallel between the two.186 Therefore, the mandalas are in fact the visual manifestation of the

conflation of the cult of relics and the cult of the book. They thus reflect a merging of devotional

practices on the painted surface that mirrored the blended religious practices of medieval Japan.

Such practices as these demonstrate the diverse lives of sacred text beyond their

discursive or hermeneutical value or as simply a medium through which the dharma is

communicated. Scriptures were valued for their materiality, their salvific, apotropaic, and

185
Miya, Kinji hōtō mandara, 7.
186
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 232-36. T no. 262, 9: 45b-46b.

172
prophylactic power, and for their sheer presence, which enlivened stūpas regardless of their

visibility. We find in religious practice the merging of sūtra and stūpa taking form in many

diverse manners, revealing again the early medieval understanding of sacred text as dharma

relics. But explanations for the imbricated central reliquary of the jeweled-stūpa mandalas have

yet to venture beyond the conclusion that the mandalas‘ are yet another incarnation of this long

tradition of combinatory practice based on the merit of both constructing stūpas and copying

sūtras in one unified project. And while this is certainly a sound and secure explanation, I believe

that the mandalas embody more than the search for the combination of multiple merits in one

manifestation, clever and practical though it is. The argument begun in this chapter and

concluded in the next offers a reading of the mandalas through what I describe as a salvific

matrix of text and body as the theoretical foundation of the paintings. The established partnership

between sūtra and stūpa becomes imbricated in the jeweled-stūpa mandalas where the concepts

of body, relic, text, and reliquary are allowed to exist in a fluid and constantly interchanging

visual relationship which, as we will see in the next chapter, reflects the definitional ambiguity

and ultimate nonduality of all things visualized in the mandalas.

Conclusion

As complex paintings with many layers of meaning, there is no single way to approach and

interpret the jeweled-stūpa mandalas. This chapter and the next work in tandem to explore the

mandalas‘ connections to Buddhist doctrine and praxis. This chapter intended to establish the

power of sacred word as perceived in early medieval Japan and demonstrate that this inherent

power was a basis for the mandalas. What is revealed through this examination about the power,

both salvific and temporal, of scriptural text uncovered persuasive reasons for the creation of the

173
mandalas. Dharma relics were shown to be nondual with the Buddha, a theme expanded in

chapter five as an integral part of the somatic and textual matrix of which the mandalas are

revealed to be the nexus. By exploring the concept of sūtra text as dharma relic manifested in

copying practices, the mandalas were shown to be not only in line with this practice but even to

have taken the notions of relic, text, body, and reliquary a step beyond through the visual

conflation of all four concepts, which ultimately reflects the underlying doctrine of nonduality. I

analyzed next the religious praxis of venerating scriptural text as relics reflected in the ubiquitous

enshrinement of sūtra in stūpas. The mandalas were revealed to be a further iteration of this well-

known and widespread practice. These lines of inquiry lay the foundation for the next chapter‘s

analysis of the mandala‘s imbricated manifestation of Buddha body theory. Ultimately, these two

chapters attempt to uncover the various and nebulous meanings of the mandalas by analyzing

their relationship to sūtra, dharma, relics, body, and stūpa, which I believe is the key to their

interpretation.

174
 Chapter Five

Jeweled-Stūpa Mandalas as a Salvific Matrix of Text and Body

Introduction

Making present the Buddha in absentia is perhaps the ultimate basis of much of the material and

visual culture of Buddhism. The parinirvāṇa, or physical death, of Siddhārtha Gautama made the

issue of absence unavoidable, sparking complex and creative ontological understandings of the

Buddha‘s nature (buddhadhātu) and raising philosophical questions about how to ‗presentize‘

the abstract, intangible, and absent. However, it is this concretized, obvious absence after the

parinirvāṇa and the need to visualize Buddhism as manifest, accessible concepts that give shape

and form to the intangible, thus greatly enriching the presence in the absence. Understanding the

Buddha in absentia necessitated theories of the Buddha‘s bodies (buddhakāya) and is manifested

visually in such things as relics, stūpas, sūtras, and icons. The jeweled-stūpa mandalas examined

in this dissertation are no exception. By presentizing the Buddha narratively, textually, and

architecturally, the mandalas suggest the many forms in which the Buddha can be made manifest

and his salvific power thus accessed.

The fourth chapter addressed the issue of the mandalas‘ privileging of sacred text as the

paintings‘ most prominent and certainly innovative feature by suggesting the power imbued in

scripture as well as the prolific practice of copying to be compelling forces ushering the

mandalas into existence. This chapter continues the discussion of the mandalas‘ reflection of

doctrine and praxis by addressing the question of the stūpa form and revealing it to be

inextricably linked to Buddha body theory. Therefore, I look at Buddha body doctrine as the

175
main unifying theory underpinning the jeweled-stūpa mandala‘s construction as the visual locus

of what I call the salvific matrix of text and body, which ultimately conflates text, dharma, body,

relic and stūpa. In order for the bodies of the Buddha to be revealed as the foundational

denominator building the mandalas, a brief discussion of the complexities and ambiguities of

Buddha body doctrine must be undertaken in order to ascertain the theory‘s relationship with the

mandalas. What is attempted is by no means a complete survey of Buddha body doctrine—which

is, as one scholar has appropriately described it, ―notorious for its complexity‖ 1—or the

discourse and debates surrounding it. It is also important to consider the place of body in the

Lotus Sūtra and the Golden Light Sūtra, the texts specifically used to construct the dharma

reliquary of the Japanese mandalas.

I also address the choice of stūpa for the visual format of the text. Building on the Lotus

Sūtra claim that the earthly body of the Buddha is the stūpa, I examine another concept of stūpa

as dharmakāya. Such an interpretation suggests further conflation of the bodies of the Buddha as

manifested in the mandalas, and once again reveals another point along the somatic strand of the

web. I argue that the identification of the stūpa as the salvifically charged, architectural body of

the Buddha—a structure housing other bodies of the Buddha in the form of ‗living‘ 2 relics, both

corporeal and dharmic—makes the monument a compelling candidate for the central icon of the

mandalas. Through these avenues of investigation I conclude that the centrality of the mandalas‘

dharma reliquaries is not a random or conceptually light choice—salvific power and multiple

iterations of body resonate in the choice of the stūpa as the iconic image of the mandalas.

The final section of the chapter analyzes the jeweled-stūpa mandalas through a matrix of

text, dharma, body, relic, and stūpa and posits that the mandalas are the visual nexus of this web.

1
Paul Harrison, ―Is the Dharma-kāya the Real ‗Phantom Body‘ of the Buddha?,‖ Journal of the International
Association of Buddhist Studies 15 no. 1 (1992): 44-94.
2
The concept of relics as living entities has been addressed in chapter four.

176
This section aims to pull together the textual and somatic strands of the salvific web explicated

throughout chapters four and five to reveal the mandalas as the visual meeting ground for these

concepts. The nuanced and ultimately intertwined concepts are all expressed in the structuring of

the jeweled-stūpa mandalas. In order to develop the assertion that this somatic and textual matrix

forms the doctrinal backbone of the mandalas, I analyze the ambiguity of the definitional

language that allows for the concepts‘ many potentialities. Doing so reveals the conflation of

their identities at play in the mandalas, as the concepts build and support as well as subsume one

another in an existential haze of nonduality at a fundamental level.

Bodies of the Buddha

Because of the prominent and rather unusual role of sacred word in the jeweled-stūpa mandalas,

the fourth chapter analyzed the inherent salvific power in scripture and introduced its layered

connection to the body of the Buddha as strong forces behind the inventive manipulation of sūtra

text within the paintings. And now, because the text of the sūtras constructs a detailed stūpa that

dominates the focus of the mandalas and because of the intertwined relationship of dharma relics

to the body of the Buddha, I continue the investigation in this chapter by further developing the

imbricated relationship of Buddha body to the mandalas.

Introduction to Buddha Body Theory

Plotting the precise development of buddhakāya doctrine from a single-bodied Buddha (as at

least one scholar argues3) to a two- and three-, or even four-bodied theory, is an impossible task

3
―[T]he earliest ideas in Mahāyāna sūtras were neither the two-body nor the three-body ones, but rather the notion
of one Buddha body.‖ Lewis R. Lancaster, ―The Oldest Mahāyāna Sūtra: Its Significance for the Study of Buddhist
Development,‖ Eastern Buddhist 8 (1975): 46. See also, Lewis R. Lancaster, ―An Early Mahāyāna Sermon about the
Body of the Buddha and the Making of Images,‖ Artibus Asiae 36 (1974): 287-91.

177
fraught with anachronistic traps. In order to discuss the foundational relationship of Buddha body

theory to the mandalas, I first introduce the fundamentals and provide a sketch of the theory‘s

trajectory over time and place. I also show how the concept is presented in the sūtra texts used in

the mandalas. In this section and the following one concerning the choice of the stūpa for the

central textual icon, while I present evidence from texts, archaeological sites, and visual culture

long predating and geographically distant from the eleventh- through thirteenth-century Japanese

context in which the mandalas were produced, this is in no way to imply the view that Buddhism

and its religious practice and material expression were monolithic. Rather my goal is to establish

the long-set precedence and earliest foundations of Buddha body theory and stūpa potency

before examining their place in medieval Japan.

As has already been discussed in the previous chapter‘s section on dharma relics, dharma

was widely considered and treated as nondual with the Buddha-nature. This early conflation of

the Buddha‘s teachings with the Buddha‘s true nature eventually established a transcendental,

and considered by many, eternal body of the Buddha identified as the dharmakāya (dharma

body; 法身 Jpn. hōshin, Ch. fashen). But an examination of the occurrences of dharmakāya in

early texts reveals that the uses of the term identified it as the ‗collection of teachings,‘ or ‗body

of teachings,‘ and as the ‗collection of dharmas‘ in which followers could seek refuge and access

to the Buddha and his law after the parinirvāṇa, rather than the highly conceptual body of the

trikāya system. Over time, scholarship on the Buddha body doctrine has corrected the tendency

in earlier studies to nominalize the early uses of dharmakāya and to ignore the plural forms of

the term, resulting in what many scholars have described as an anachronistic reading of

dharmakāya as the fully developed transcendental body corresponding to the later trikāya theory,

effectively mischaracterizing the development of the doctrine as far too consistent and tidy. Paul

178
Harrison, through extensive research on Buddha body doctrine, concludes that many of the early

uses of dharmakāya should be translated as ‗body of dharmas‘ rather than the more specific and

loaded term ‗dharmabody.‘4 He determines that rather than establishing a distinct spiritual body

of the Buddha, the main intention in these early texts is to equate the Buddha with the dharma,

and that even when dharmakāya occurs in the nominal case it refers to the body of scriptures.

Thus he suggests the emphasis be placed on the dharma rather than the kāya. Most scholars on

the subject of Buddha bodies have reached similar conclusions about the early understandings of

the bodies of the Buddha.5

Such is the case in the Nirvāṇa Sūtra, the Pali texts, Sūtra on the Questions of King

Miliṇḍa and Kindred Sayings,6 the Perfection of Wisdom Scripture in Eight Thousand Lines, and

the Diamond Sūtra, among numerous others.7 An emblematic example comes from the

Perfection of Wisdom Scripture in Eight Thousand Lines: ―Indeed, the Tathāgata is not to be seen

in the body of form and shape [rūpakāyato]. For the bodies of the teachings [dharmakāyāḥ]—

4
Paul Harrison, ―Is the Dharma-kāya the Real ‗Phantom Body‘ of the Buddha?.‖
5
For example, see Paul Demiéville, ―Busshin,‖ in Hōbōgirin, ed. Paul Demiéville (Tokyo: Maison-Franco-
Japonaise, 1929), 174-85; Eckel, To See the Buddha; Guang Xing, The Concept of the Buddha: Its Evolution from
Early Buddhism to the Trikaya Theory (London and New York: Routledge Curzon, 2005); Ruben L. F. Habito, ―The
Notion of Dharmakāya: A Study in the Buddhist Absolute,‖ Journal of Dharma 11 (1986): 348-378, ―Buddha-body
Theory and the Lotus Sutra: Implications for Praxis,‖ in A Buddhist Kaleidoscope: Essays on the Lotus Sutra, ed.
Gene Reeves (Tokyo: Kōsei Publishing Company, 2002), 305-18; Yuichi Kajiyama, ―Stūpas, the Mother of
Buddhas, and Dharma-body,‖ in New Paths in Buddhist Research, ed. A. K. Warder (Durham, North Carolina: The
Acorn Press, 1985), 9-16; Lancaster, ―An Early Mahayana Sermon about the Body of the Buddha and Making of
Image;‖ Whalen W. Lai, ―The Predocetic ‗Finite Buddhakāya‘ in the Lotus Sūtra: In Search of the Illusive
Dharmakāya Therein,‖ The Journal of the American Academy of Religion XLIX no. 3 (1981): 447-469; Nagao
Gadjin, ―On the Theory of Buddha-body: Buddha-kāya,‖ trans. Hirano Umeyo, Eastern Buddhist 6 (1973): 25-53;
John J. Makransky, "Controversy over Dharmakāya in India and Tibet: A Reappraisal of Its Basis,
Abhisamayālaṃkāra Chapter 8," The Journal of the International Association of Buddhist Studies 12 (1989): 45-78,
Buddhahood Embodied: Sources of Controversy in India and Tibet (Albany, NY: State University of New York
Press, 1997); and Robert H. Sharf, Coming to Terms with Chinese Buddhism: A Reading of the Treasure Store
House (Honolulu: University of Hawai‘i, 2002), 100-11. This is not to suggest a uniformity of opinion on the
Buddha body doctrine among these scholars.
6
Jpn. Zō agon kyō; Ch. Za ahan jing; Skt. Saṃyutta nikāya; 雜阿含経; T. no. 99, 2: 1a-373b.
7
For passages from these texts supporting the early understanding of dharmakāya, see chapter four‘s section on
dharma relics.

179
these are Tathāgatas.‖8 This passage also typifies the sentiment prevalent in descriptions of the

dharmakāya, namely one of bodily bifurcation: the division from—and often privileging of—the

dharmakāya over the earthly form or rūpakāya (色身 Jpn. shikishin, Ch. seshen) assumed by

Śākyamuni during life on earth. Because the rūpakāya is ultimately only a manifestation of the

Buddha, and as such cannot embody the dharmatā (the all-pervading truth) within the realm of

form, only the dharmakāya is synonymous with the pure essence of buddhahood and truth. The

rūpakāya was portrayed as fundamentally tainted as an emanation in the realm of form, but at the

same time was celebrated as the most beautiful manifestation, as Bhāvaviveka, a sixth-century

Madhyamakan (Middle Way tradition; 中觀派, Jpn. Chugan ha, Ch. Zhongguan pai) monk

describes in his work, The Verses on the Essence of the Middle Way:9

The [Buddha‘s] incomparable Form [Body] is surrounded by a fathom of light that has
the appearance of a rainbow; its splendor consists of permanent, radiant, and complete
primary and secondary characteristics; its ornament is glory; it is charming to the mind
and eyes; and it surpasses all things in beauty.

[With this form] and with a miraculous voice that has sixty attributes [the Buddha]
captivates the minds of all beings.

With body and voice like a wishing jewel, [the Buddha] assumes the universal form of all
the gods to help those who are ready to be taught.10

There is broad consensus that the works of Nāgārjuna11 and the Mādhyamaka tradition

presented and popularized a thoroughly bifurcated image of the bodies of the Buddha: that of the

dharmakāya and rūpakāya. As Malcolm David Eckel has pointed out, this two-body system

grew out the canonical distinction of the vile body of form and the body of pure teachings or

8
Habito, ―Buddha-body Theory and the Lotus Sutra,‖ 306.
9
Jpn. Daijō shōchin ron; Ch. Dasheng zhangzhen lun; Skt. Madhyamakahṛdayakārikās; 大乘掌珍論; T. no.1578,
30: 268a26-278b09.
10
Eckel, To See the Buddha, 124-25.
11
For a cursory look at Nāgārjuna‘s view of Buddha body, see Nagata Mizu 永田瑞, ―Ryūju no busshinkan 竜樹の
仏身観,‖ Indogaku bukkyōgaku kenkyū 印度学仏教学研究 22, no. 1 (1973): 369-72.

180
dharmas as seen in the Perfection of Wisdom Sūtras and the Lotus Sūtra.12 The Mahāyāna

commentator, Mātṛceṭa (first century CE), reveals the two-body system in his work, One

Hundred and Fifty Verses:13 ―Even when you had attained nirvana, you showed the unbelieving

world, ‗My Dharma and Form Bodies are meant for others.‘ For when you handed over the

Dharma Body completely to the virtuous and split the Form Body into parts, you attained

parinirvana.‖14 Nāgārjuna‘s Jeweled Garland15 offers a similar sentiment: ―If the causes of the

Buddha‘s Form Body are as immeasurable as the world, how can someone measure the cause of

the Dharma Body? The Buddha‘s Form Bodies arise from the collection of merit; and, to put the

matter briefly, O King, the Dharma Body is born from the collection of insight.‖16 Much in this

way, the Mādhyamaka tradition continued to popularize the two-body system of the Buddha

which was later adapted into what became the more standard view of the Buddha‘s bodies: the

three-body theory.

The trikāya doctrine as propagated by the Yogācāra-vijñāna school offered the following

somatic scheme: svābhāvikakāya (essence body 自性身), sāṃbhogikakāya (enjoyment body 受

用身), and nairmāṇikakāya (transformation body 變化身).17 As John J. Makransky points out,

the sāṃbhogikakāya refers to the illustrious tathāgatas found in the Mahāyāna sūtras, and the

nairmāṇikakāya corresponds to the innumerable manifestations of the Buddha in the world of the

12
Eckel, To See the Buddha, 115. For canonical quotes supporting this distinction, see the previous chapter‘s
section on dharma relics.
13
Jpn. Ippyakugojū sanbutsuju; Ch. Yibaiwushi zanfosong; Skt. Śatapañcāśatka; 一百五十讚佛頌; T. no. 1680, 32:
758b23-762a13.
14
Eckel, To See the Buddha, 115. An alternate translation is available in D.R. Shackleton Bailey, ed., The
Śatapañcāśatka of Mātṛceṭa (Cambridge: Syndics of the Cambridge University Press, 1951), 179. Śatapañcāśatka
144b-45.
15
Jpn. Hōgyō ōshō ron; Ch. Baoxing wangzheng lun; Skt. Ratnāvalī; 宝行王正論; T. no. 1656, 32: 493b4-505b1.
16
See Eckel, To See the Buddha, 115; Ratnāvalī III.10 and 12. For another translation, see Jeffrey Hopkins and Lati
Rimpoche with Anne Klein, eds. and trans., The Precious Garland and The Song of the Four Mindfulnesses:
Nāgārjuna and the Seventh Dalai Lama (London: George Allen and Unwin Ltd., 1975) verses 210-212, p. 48-49.
17
Nagao, ―On the Theory of Buddha-body,‖ 30-39.

181
unenlightened. 18 These terms roughly correlate to the later terms, which succeeded them in

popular usage: the saṃbhogakāya (reward, enjoyment or retribution body; 報身 Jpn. hōjin, Ch.

baoshen) and nirmāṇakāya (manifestation or transformation body; 化身 Jpn. keshin, Ch.

huashen).19 According to the Sūtra on (the Buddha's) Entering (the Country of) Laṅkā,20 the

saṃbhogakāya is the body of the Buddha that experiences enlightenment in the pure palace of

Akaniṣṭha while it is only a manifestation (nirmāṇakāya) that is enlightened in the polluted

world of the living.21 The saṃbhogakāya, while still categorized in the realm of form along with

the nirmāṇakāya, is only visible to those of enlightened capabilities for the purposes of shared

enjoyment of the dharma. The saṃbhogakāya is described as possessing the thirty-two lakṣaṇa.

The nirmāṇakāya is the perishable body with which the Buddha in his great compassion causes

infinite emanations to manifest in the unenlightened realm of the living. According to the quasi-

mythical Yogācāra scholar, Maitreyanātha, in his commentary on the Perfection of Wisdom

Sūtras entitled Ornament of Clear Realization,22 ―[The body] by which he brings various

benefits to living beings without interruption as long as there is samsara is the Manifestation

Body of the Sage.‖23 Śākyamuni Buddha was one such manifestation.

The svābhāvikakāya roughly corresponds to the dharmakāya of the trikāya system. As

Makransky explains, the svābhāvikakāya in early Yogācāra literature is understood as ―being the

dharmakāya, whose character is aśrayaparāvṛtti [Yogācāran characterization of buddhahood as

18
Makransky, "Controversy over Dharmakāya in India and Tibet,‖ 53.
19
Aside from the most commonly used terms, a discussion of the multitude of names designating the bodies of the
Buddha cannot be undertaken here. For a very brief summary of the various names, see Nagao, ―On the Theory of
Buddha-body: Buddha-kāya,‖ in particular footnote 6.
20
Jpn. Ryōgakyō; Ch. Lengjiajing; Skt. Laṅkāvatāra sūtra; 楞伽経; T. no. 671, 16: 514c4-586b22.
21
Eckel, To See the Buddha, 126.
22
Jpn. Genkan shōgon ron; Ch. Xianguan zhuangyan lun; Skt. Abhisamayālaṃkāra śāstra; 現觀莊嚴論.
23
Eckel, To See the Buddha, 216 n61.

182
a shift in standpoint or basis].‖ 24 ―In other words, they equate svābhāvikakāya with dharmakāya

in its sense of buddhahood as a whole. But why, one might ask, do we need another term for all

of buddhahood? We already have so many of these terms. The answer is that there is

buddhahood as it actually exists, i.e., as a buddha has realized it (svābhāvikakāya).‖25 Makransky

explains, ―A buddha has achieved only one buddhahood, the dharmakāya.‖26 He also notes that

svābhāvikakāya since its usage in sūtras has been a controversial term, not fully understood or at

least functioning in subtly different ways depending on the text, a controversy that is

communicated in commentaries and in scholarship today. 27 Gradually, however, dharmakāya

came to replace svābhāvikakāya once influential commentators such as Vasubandhu, Asvabhāva,

and Sthiramati lent it greater prominence by adopting it in their commentarial texts. 28

The dharmakāya of the three-body theory is one of transcendent, undefiled essence, such

as that described by Jñānagarbha, an eighth-century Mādhyamika monk:

When [the Buddha] takes no notice of subject, object, or self, no signs of cognition arise
[in his mind]. His concentration is firm, and he does not get up.

The place where he sits is a locus (sthāna) of every inconceivable virtue. It is

incomparable, worthy of worship, a guide, and utterly beyond thought.

This is the Dharma Body of the Buddhas, because it is the body of all the qualities
(dharma) [that constitutes a Buddha], the locus (aśraya) of every inconceivable virtue,
and rational in nature.29

While still denoting the teachings, and hence the scriptures as evidenced in the praxis of sūtra

worship discussed in the last chapter, dharmakāya acquires a more abstract and grand

philosophic dimension. In his examination of texts such as the Jewel Nature Treatise,30 an

24
Makransky, "Controversy over Dharmakāya in India and Tibet,‖ 55.
25
Ibid.
26
Ibid.
27
Ibid., 48-51.
28
Ibid., 68-69 n35.
29
Eckel, To See the Buddha, 65; whose translation is adapted from his own Jñānagarbha Commentary.
30
Jpn. Hōshōron; Ch. Baoxinglun; Skt. Ratnagotravibhāga; 宝性論; T. no. 1611, 31: 813a8-48a27.

183
important work in tathāgatagarbha thought which holds that all beings possess the ability to

achieve buddhahood, Ruben L. F. Habito characterizes the dharmakāya as the all-pervading

principle of unity and truth by which all living beings are encompassed, highlighting several

passages from the sūtra, such as, ―The universal body (dharmakāya) of the Tathāgata penetrates

all living beings.‖31 The dharmakāya is truth without cessation, it is ―not a body, is not created,

is not born, does not cease, is not produced by a combination [of causes], does not arise, does not

remain, is not established, has no end, has no limit, is happy and is utterly quiet.‖32 It is from the

dharmakāya that the other bodies emanate. The dharmakāya is broadly understood to be the

Buddha‘s teachings, the ultimate truth embodied in those lessons, as well as the realization of

that truth; thus, we have a body composed of sūtra, embodying the ultimate truth by which the

realization of such truth is attained, effectively making the dharmakāya both the path and the

goal.

Buddha Body in the Lotus Sūtra and Golden Light Sūtra

While the Lotus Sūtra and Golden Light Sūtra do not present a systematic, fully fledged vision of

the trikāya doctrine, important for the mandalas, they do offer a view of the eternal Buddha

accessible though the dharma. In chapter sixteen of the Lotus Sūtra, ―The Life Span of the Thus

Come One,‖33 the Buddha confesses that even though he presently lives the life Śākyamuni—a

prince of the Śākya clan who forsook earthly pursuits and pleasures, thus attaining

anuttarasamyaksaṃbodhi (the utmost and perfect enlightenment; 阿耨多羅三藐三菩提 Jpn.

31
Habito, ―The Notion of Dharmakāya,‖ 357.
32
Eckel, To See the Buddha, 166. Bhāvaviveka is quoting the Tathāgatajñānamudrāsamādhi sūtra in his
commentarial work, Tarkajvālā (The Flame of Reason).
33
Jpn. Nyorai juryo bon; Ch. Rulai shouliang pin; 如來壽量品.

184
anokutara sanmyaku sanbodai, Ch. anouduolo sanmiao sanputi) at Gayā—in fact, he was

enlightened in the beginningless, incalculable past:

And yet, O good men, since in fact I achieved buddhahood it has been incalculable,
limitless hundreds of thousands of myriads of nayutas of kalpas. For example, one might
imagine that in the five hundred thousand myriads of millions of nayutas of asaṃkhyeyas
of thousand-millionfold worlds there is a man who pounds them all to atoms, and then,
only after passing eastward over five hundred thousand myriads of millions of nayutas of
asaṃkhyeyas of realms, deposits one atom, in this way in this eastward movement
exhausting all these atoms. 34

Indeed, the Buddha reveals that for ―a hundred thousand myriads of millions of nayutas of

asaṃkhyeyas I have been constantly dwelling in this Sahā world sphere, preaching the dharma,

teaching and converting; also elsewhere, in a hundred thousand myriads of millions of nayutas of

asaṃkhyeyas of realms [I have been] guiding and benefiting the beings.‖ 35 Thus, the Lotus Sūtra

portrays a Buddha, limitless and eternal, continuously manifesting the dharma via the

compassionate method of expedient or skilful means. Moreover, the Lotus Sūtra characterizes

other Buddhas from other realms as emanations of the body of Śākyamuni, 36 thereby subsuming

all Buddhas under the one ceaseless, limitless Buddha presently identified as Śākyamuni.

As examined in the last chapter, the Lotus Sūtra explicitly and unequivocally equates

itself with the body of the Buddha, saying, ―O Medicine King! Wherever it may be preached, or

read, or recited, or written, or whatever place a roll of this scripture may occupy, in all those

places once is to erect a stūpa of the seven jewels, building it high and wide with impressive

decoration. There is no need even to lodge śarīra in it, what is the reason? Within it there is

already a whole body of the Thus Come One.‖37 In another passage, the scripture reaffirms the

nonduality of the sūtra and the Buddha‘s body: ―If there is anyone who can hold it [the Lotus

34
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 219. T. no. 262, 9: 42b12-16.
35
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 220. T. no. 262, 9: 42b25-28.
36
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 170, 211, ―Apparition of the Jeweled Stūpa‖ and
―Welling Up Out of the Earth,‖ respectively. T. no. 262, 9: 32c22-27 and 41a4-6.
37
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 163. T. no. 262, 9: 31b26-29.

185
Sūtra], / Then he holds the buddha body.‖ 38 And critical to the multiple visions of Buddha body

in the mandalas is the equivalence of the Buddha‘s body with the stūpa, as testified to in the

Lotus Sūtra, a topic I examine in the next section.

The Golden Light Sūtra also speaks of the incalculability of the Buddha‘s existence. In

the second chapter, ―Measure of Life of the Tathāgata,‖ the assembled Tathāgatas proclaim in

united voice:

The drops in all the oceans of water can be counted, but no one can count the life of
Śākyamuni. As far as the Sumeru mountains are concerned, all their atoms can be
counted, but no one can count the life of Śākyamuni. However many atoms there are on
earth it is possible to count them all but not to count the life of the Buddha. If anyone
should wish to measure the sky, (it is possible), but no one can count the life of
Śākyamuni. Let there be some many æons and hundreds of millions of æons, so many
perfect Buddhas, yet the count (of his life) is not obtained. 39

In the same chapter, the brahmin, Kauṇḍinya, describes that Śākyamuni ―is not created and has

not arisen. His body that is as hard as the thunderbolt manifests his transformed body. And hence

there is nothing called a relic of the great sage even the size of a grain of mustard. How will there

be a relic in a body without bone or blood? The depositing of a relic is by an expedient on

account of the welfare of beings.‖40 Kauṇḍinya explains, ―For the one who has the Law as his

body is the one fully enlightened; the sphere of the Law is the Tathāgata. Such is the Lord‘s

body; such the exposition of the Law.‖41 In unison, the congregation of thirty-two thousand gods

exclaim: ―The Buddha does not enter complete Nirvāṇa (and) the Law does not disappear. For

the ripening of beings does he teach complete Nirvāṇa. The Lord Buddha is inconceivable. The

Tathāgata has an eternal body. He shows various manifestations by reason of the welfare of

38
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 176. T. no. 262, 9: 34b12.
39
Emmerick, The Sūtra of Golden Light, 5. T. no. 665, 16: 405a13-18.
40
Emmerick, The Sūtra of Golden Light, 7-8. T. no. 665, 16: 406c08-12.
41
Emmerick, The Sūtra of Golden Light, 8. T. no. 665, 16: 406c13-14.

186
beings.‖42 Again, similar themes involving a timeless, compassionate Buddha manifesting the

dharma through upāya, formless and nondual with the Law appear in the Golden Light Sūtra.

Such a Buddha thus emanates as the ground or source all other manifestations rendered in form.

In the jeweled-stūpa mandalas, body is visualized in several forms. The nondual

conflation of dharma as Buddha (as explored in the last chapter), along with the concept of

dharmakāya–as body of scripture and as the body of ultimate truth encapsulated in the law—are

all revealed to be strands leading to the culmination of the mandalas as the nexus of a textual and

somatic web. The full implications of the textual and somatic layers visualized in the dharma

reliquary are excavated in the third section of this chapter. In the next section, analyzing the

choice of the stūpa for the textual icon demonstrates that the bodily references manifested in the

jeweled-stūpa mandalas are further imbricated.

Choosing the Stūpa

The question compelling this section—namely, the choice of the stūpa for the textual icon—

observes the corporeal dimension and worship of the reliquary, the notion of the stūpa as a

salvific monument, and the prolific commissions of stūpa projects in early medieval Japan. As a

ubiquitous monument in Buddhist realms, I aim to examine the significance of the stūpa and

address its place in the mandalas. This three-pronged approach into the choice of the stūpa

further explicates the strands of the web dealing with the somaticity of the Buddha and the

salvific spheres accessible in our realm, all of which highlight the mandalas as the karmic

convergence of these concepts.

42
Emmerick, The Sūtra of Golden Light, 8. T. no. 665, 16: 406c18-21. The Chinese translation by Yijing contains a
chapter on the trikāya doctrine not included in the Sanskrit versions and Emmerick‘s translation, called
"Distinguishing the Three Bodies" (分別三身品 Jpn. Funbetsu sanjinbon, Ch. Fenbie sanshenpin).

187
Corporeality of the Stūpa

In an inventive take on the transformative role of sacred word, Charlotte Eubanks comments that

in ‗writing‘ sacred text onto memory during medieval Buddhist practice, sūtras become

internalized.43 As she explains, ―The sūtra inside, created through an amalgamation of sensory

experiences and textual encounters, is an embodied one in the sense that it is surrounded by and

housed in the body, creating physically legible symptoms….‖ 44 Such internalization transforms

the body into a jeweled-stūpa (宝塔 Jpn. hōtō, Ch. baota).45 Thus body can be conceptualized

and sacralized in a multitude of ways, and there is no one definitive way to think of the notion of

body. True to this, the mandalas do not exhibit just one definition of Buddha body; they exhibit

several. In this section, I continue the somatic thread, building on the earlier explications of

sacred body and the layered bodily visualizations in the jeweled-stūpa mandalas by revealing the

stūpa to be another manifestation of the body.

As John S. Strong has noted, the ―apparent functional equivalence of stūpa and buddha‖

in the worship of stūpas stems from the conviction that ―a stūpa ‗is‘ the living buddha.‖ 46 Both

doctrine and praxis confirm this corporealization of the stūpa. For example, in chapter eleven of

the Lotus Sūtra, ―Apparition of the Jeweled Stūpa,‖ the Buddha instructs his disciple in the

proper post-parinirvāṇa methods of veneration saying, ―After my passage into extinction,

anyone who wishes to make offerings to my whole body must erect a great stūpa.‖ 47 As is

discussed later in the chapter, the Lotus Sūtra adeptly straddles the devotional line between the

cult of the book and the cult of the stūpa, a synthesis of extremes that in all likelihood reflected

the blended practice of medieval Buddhism. Therefore, when the Buddha unequivocally equates

43
Eubanks, ―Rendering the Body Buddhist‖, 305.
44
Ibid.
45
Ibid., 306.
46
John S. Strong, Relics of the Buddha (Princeton: Princeton University Press, 2004), 32.
47
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 168. T. no. 262, 9: 32c15-16.

188
his ‗complete body‘ to that of a grand architectural monument, i.e. a stūpa, any distinction

between a stūpa and the Buddha is abolished. The Buddha, presciently perhaps, articulates a

strategy for manifesting his presence in absentia, thus elevating the status of the stūpa to the

nondual level with the Buddha.

Several studies have been conducted comparing the architectural structure of the stūpa

with the attributes of the dharmakāya as detailed in sūtras. Overall, these studies have attempted

to characterize the presence of the Buddha‘s seeming absence. Gustav Roth argues that the stūpa

is equivalent to the Buddha by analyzing passages from the Caitya vinayodbhāva sūtra in its

Tibetan form, the Sansrkit manuscript Stūpa lakṣaṇa kārikā vivecana, and the

Kriyāsaṃgrahapañjikā of Kuladatta, a Buddhist Tantric ritual compendium, among other

sources.48 Quoting from the Caitya vinayodbhāva sūtra on the nature of the dharmakāya, Roth

writes, ―The substance of the Dharmakāya are the applications of mindfulness, the exertions, the

moral faculties, the abilities….‖49 As he notes, the sūtra then reveals that the stūpa embodies

these enumerated elements: ―The stūpa is the reflected image of these … i.e. the reflected image

of the Dharmakāya.‖50 The texts examined by Roth dissect each part of the stūpa, equating the

sections with various characteristics of the Buddha‘s essence, thereby exposing the stūpa as yet

another manifestation of the Buddha‘s body, and more specifically, the dharmakāya of the

Buddha. Indeed, from the foundational terrace steps to the crowning canopy, the characteristics

of the dharmakāya embody each section of the stūpa structure, articulating in architectural form

a terrestrial dharmakāya in absentia. For example, the arguably dominant feature of the stūpa,

48
Gustav Roth, ―Symbolism of the Buddhist Stūpa according to the Tibetan version of the Caitya-vibhāga-
vinayodbhāva-sūtra, the Sanskrit treatise Stūpa-lakṣaṇa-kārikā-vivecana, and a corresponding passage in Kuladattas
Kriyāsaṃgraha,‖ in The Stūpa. Its Religious, Historical and Architectural Significance, ed. Anna Libera
Dallapiccola in collaboration with Stephanie Zingel-Avé Lallemant (Wiesbaden: Steiner, 1980), 183-209.
49
Ibid., 187.
50
Ibid.

189
the kumbha (the dome or ‗pot‘ as Roth calls it), has been shown to have roots in the rituals of

relic preservation. Both the Sanskrit Nirvāṇa Sūtra and the Pāli Mahāparinibbāna sutta describe

the collection of the Buddha‘s ashes into the kumbha as a relic repository for the Tathāgata‘s

corporeal remains. 51 In terms of the stūpa‘s symbolic equivalent of the dharmakāya’s essence,

the kumbha embodies the seven constituents of enlightenment.52 As Peter Harvey as summarized,

―the stūpa, the primary focus of early Buddhist development, should not only contain the relics

of the Buddha or a saint, but should also symbolise the Dhamma, or the Buddha in the form of

his Dhammakāya.‖53

Religious practice reflects this corporeality of the stūpa, as Gregory Schopen has

persuasively argued in his study on the medieval practice of burial ad sanctos (a Latin verse

meaning ‗at the place of saints‘).54 Schopen exposes the ―functional equivalence of the relic and

the living buddha‖55 in his examination of inscriptions from early Indian reliquaries, ca. first and

second century CE. Two very early inscriptions from the lid of a reliquary during the reign of

King Menander (second century BCE) describe Śākyamuni‘s relics as having life: ―…[on] the

14th day of the month Kārttika, the relic of the blessed one Śākyamuni which is endowed with

life is established,‖ and ―[This is] a relic of the blessed on Śākyamuni which is endowed with

life.‖56 Such testimonials reveal relics as living entities. But relics of the Buddha are shown to

embody more than just life—inscriptions from the first and second century CE describe relics as

51
Peter Harvey, "The Symbolism of the Early Stūpa," Journal of the International Association of Buddhist Studies 7
no. 2 (1984): 71. Harvey offers several interesting symbolic interpretations of the stūpa form; however, his
arguments would have been more persuasive if more space had been provided in order to fully develop his
assertions.
52
According to The Pali Text Society’s Pali-English Dictionary, the seven constituents of enlightenment
(bojjhaṅgā) are: ―sati, dhamma vicaya, viriya, pīti, passaddhi, samādhi, and upekhā or mindfulness, investigation of
the Law, energy, rapture, repose, concentration and equanimity.‖
53
Harvey, "The Symbolism of the Early Stūpa,‖ 84.
54
Gregory Schopen, ―Burial Ad Sanctos and the Physical Presence of the Buddha in Early Indian Buddhism,‖
Religion 17 (1987): 193-225.
55
Ibid., 209.
56
Ibid., 204.

190
possessing the attributes of a living Buddha, such characteristics as wisdom, vision, morality,

virtue, and emancipation.57 Thus, relics were perceived to be impregnated with the dharma and

essence of the Buddha, making them not just ‗like‘ the Buddha, but nondual entities.

Stūpas, enshrining corporeal and dharmic relics—in both of which the Buddha is

present—were thus viewed as architectural bodies of the Buddha. According to the inscriptions

from stūpa number one at Sāñcī, ca. first century BCE, to do harm through desecration or theft to

a stūpa was a harshly punishable sin: ―He who dismantles, or causes to be dismantled, the stone

work from this Kākaṇāva [i.e. the old name for the stūpa at Sāñcī], or causes it to be transferred

to another ‗house of the teacher‘, he shall go to the [same terrible] state as those who commit the

five deadly sins that have immediate retribution.‖ 58 As Schopen notes, this inscription is telling.

Because committing crimes against the stūpa is tantamount to the five deadly sins (matricide,

patricide, killing an arhat, causing divisions within the sangha, and physically harming a

Buddha), for such grievous sins to apply means that a stūpa was viewed as a ‗living person of

rank.‘ That desecration and the removal of offerings made to the stūpa were punishable sins

indicates a legal status accorded to the stūpa much as that of a ‗legal person,‘ with all the

accompanying rights, privileges, and protections. Schopen reveals through an investigation of the

legal and karmic protection afforded stūpas in the monastic codes (vinayas) and Mahāyāna sūtras

that early stūpas were indeed conceived of as a ‗legal person‘ capable of owning property,

including land and monetary funds. For example, the Ratnarāśi sūtra59 declares that all money

and objects given to a stūpa cannot be used by the populace or clergy or exposed to the elements,

they must even be allowed to rot. The Ratnarāśi sūtra explains that the reason for such

protectionist measures is that the stūpa as the Buddha sacralizes these objects through possession:

57
Ibid., 204-06.
58
Ibid., 206-07.
59
Jpn. Hōjūkyō; Ch. Baojujjing; 宝聚経.

191
―Whatever belongs to a stūpa, even if it is only a single fringe that is given … that itself is a

sacred object for the world together with its gods.‖60 These examples highlight the early tradition

of stūpa veneration and identification as a body of the Buddha.

The stūpa as a body of the Buddha occurs as well in visual culture. Before analyzing its

expression in the jeweled-stūpa mandalas, it is instructive to look at a few of the earliest

Buddhist images featuring the worship of a stūpa as a Buddha to demonstrate the stūpa‘s long-

standing as a reverential object. A white sandstone relief carving from the first century CE on the

northern gateway pillar of Sāñcī‘s stūpa number one, topically titled The Malla Nobles Rejoicing

on Receiving Their Share of Śākyamuni’s Relics,61 portrays a scene of stūpa worship.62 In the

horizontally arranged relief, the Mallas of Kuśinagara venerate the stūpa through dance, song,

music, feasts, and offerings of flower wreaths. Kiṃnaras in flight (half-human, half-bird gods

associated with celestial music) lay flower garlands upon the stūpa. Other such visualizations

come from the remains of the Bhārhut and Amarāvatī stūpas. The second-century BCE Bhārhut

relief, Veneration of the Stūpa, comes from the pillar featuring the religious rituals of King

Prasenajit, a ruler contemporary with Śākyamuni. 63 The red sandstone relief describes the

virtuous pilgrimage of King Prasenajit 波斯匿王 and his consort to a stūpa, perhaps even the

stūpa erected at the site of the Buddha‘s parinirvāṇa as suggested by the two śāla trees. The king

and his consort are pictorialized, in continuous narration, approaching the stūpa, bowing before it,

and circumambulating its perimeter while winged beings fly overhead offering gifts of flowered

garlands. The second-century CE Amarāvatī relief, Stūpa Venerated by Elephants, depicts in

60
Schopen, ―Burial Ad Sanctos,‖ 208.
61
For an image, see David L. Snellgrove, ed., The Image of the Buddha (London: Serindia Publications/UNESCO,
1978), 31 fig. 11.
62
For a study on the ‗aniconic‘ argument of early Buddhist imagery, see Susan L. Huntington, ―Early Buddhist Art
and the Theory of Aniconism,‖ Art Journal 49 (1990): 401-408, ―Aniconism and the Multivalence of Emblems:
Another Look,‖ Ars Orientalis 22 (1992): 111-56.
63
For an image, see Snellgrove, The Image of the Buddha, 32 fig. 12.

192
white limestone grand elephants with botanical offerings grasped in their trunk kneeling before

the stūpa in worship. 64 Nāgas (serpent deities; 龍 Jpn. ryū, Ch. long), often guardians of the

relics of the Buddha, intertwine and knot around the stūpa. In all three reliefs depicting the

worship of a stūpa in the manner befitting the Buddha, the surface of the stūpa is already adorned

with flower arrangements and/or elaborate designs.

As described in the last chapter, such manner of veneration accords with the proper

worship of the Buddha as instructed in doctrine and indicates stūpas were revered as the Buddha.

For example, in the Mahāparinibbāna sutta of the Dīghanikāya (‗Collection of Early or Long

Discourses‘), great benefits are awarded those who pay proper veneration of the stūpa: ―At the

four cross roads a cairn [stūpa] should be erected to the Tathāgata. And whosoever shall there

place garlands or perfumes or paint, or make salutation there, or become in its presence calm in

heart—that shall long be to them for a profit and a joy.‖ 65 The eighth-century Indian monk,

Śāntideva, devotes an entire chapter to stūpa worship in his Training Anthology.66 Quoting from

the Avalokana sūtra, Śāntideva urges the worship of stūpas because of the tremendous,

unparalleled merit gained. He reveals that in merely offering a garland to a stūpa, one would

become ―an imperial monarch and Śakra the lord, and Brahma in Brahma‘s world.‖ 67

Furthermore, ―…he who gives one sunshade, adorned and brilliant to see, to the Blessed One‘s

shrines‖ is awarded a body resplendent and perfect as the Buddha‘s and celebrated and

worshiped as a most virtuous and all-knowing being.68 Many more texts and sūtras command the

64
For an image, see Ibid., 32 fig. 13.
65
T.W. and C.A.F. Rhys Davids, trans., Dialogues of the Buddha Translated from the Pali of the Dīgha Nikāya,
Part II (London: Henry Frowde, Oxford University Press, 1938), 156. Dīghanikāya, II, 142.
66
Cecil Bendall and W.H.D. Rouse, trans., Śikshā-samuccaya: A Compendium of Buddhist Doctrine (London: John
Murray, 1971[1902]), 270-82. Jpn. Daijō shū bosatsu gaku ron; Ch. Dasheng ji pusa xue lun; Skt. Śikṣāsamuccaya;
大乗集菩薩学論; T. no. 1636, 32: 75b4-145a29.
67
Bendall and Rouse, Śikshā-samuccaya, 271.
68
Ibid., 272.

193
worship of stūpas. At multiple points in the Lotus Sūtra, good people are encouraged to worship

stūpas; for instance, having erected a grand stūpa to house the Lotus Sūtra as the dharma relic,

the Buddha offers these instructions: ―This stūpa is to be showered with offerings, humbly

venerated, held in solemn esteem, and praised with all manner of flowers, scents, necklaces, silk

banners and canopies, music skillfully sung and played, if there are persons who can see this

stūpa and worship and make offerings to it, be it known that these persons are all close to

anuttarasamyaksaṃbodhi.‖69 Due to the prominence of stūpa veneration in the Lotus Sūtra,

jeweled-stūpas became objects of worship themselves. Brian Ruppert, in his illuminating study

on the many uses of relics in medieval Japan, and in particular the harnessing of their power for

authoritative purposes, notes that in the Buddhist Relic Offerings rite in which relics are offered

to shrines throughout the country, a jeweled-stūpa served as the main icon, thereby expressing

the ―fecundity of the body of the buddha.‖ 70

Images such as the 1203 wooden gorintō (five-ringed stūpa) at Shin-Daibutsuji 新大仏寺

in Mie prefecture constructed of one thousand miniature seated Buddhas further defines the stūpa

as equivalent with the Buddha. 71 Again, in a combinatory instance of sūtra and stūpa, the back of

the gorintō features the text of Dhāraṇī of the Seal on the Casket written in Siddhaṃ-style letters

(梵字 Jpn. bonji, Ch. fanzi).72 This scripture, examined in the previous chapter, clearly conflates

sacred word with the bodies of the Buddhas of the past, present, and future. Thus, the gorintō

manifests the bodies of the Buddha in the form of stūpa and sūtra. A strong tradition in Shingon

Buddhism holds that the stūpa, and in particular the gorintō, is the architectural manifestation of

Dainichi as the dharmakāya. This gorintō reveals the stūpa more specifically to be Dainichi
69
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 163. T. no. 262, 9: 31b29-31c03.
70
Ruppert, Jewel in the Ashes, 68.
71
The stūpa measures 60 cm x 27 cm. For an image, see Ishida Mosaku, Nihon buttō, zuhan 日本佛塔, vol. 2
(Tokyo: Kōdansha, 1969), fig. 664.
72
Ibid., 239.

194
because of the visualization of one thousand emanating seated Buddhas, the iconography of

Dainichi grounded in doctrine. The Brahma Net Sūtra describes the emanation of one thousand

Śākyamuni Buddhas from Dainichi as the dharmakāya and thus ground or origin of all things,

including the Buddhas. The twelfth-century copy of the Chinese original brought to Japan by

Kūkai of the iconographic drawing for the Kongōkai mandara 金剛界大曼荼羅, or ‗diamond-

world mandala,‘ visualizes the stūpa as the dharmakāya of Dainichi. Inscribed on one of the

supporting lotus petals is the Sanskrit letter, ‗vaṃ,‘ for Dainichi of the diamond realm and the

stylized, adamantine thunderbolt (金剛 Jpn. kongō, Ch. jingang; Skt. vajra), another symbol

denoting Dainichi, rests at the base of the stūpa.

The stūpa as the architectural expression of Buddha body and the sūtras as the textual

body of the Buddha render visible the dharmakāya in the jeweled-stūpa mandalas. This rare

combinatory visualization presents the two manifestations of the dharmakāya that are possible to

witness in our post-parinirvāṇa realm. Examining the theories of the bodies of the Buddha as the

key to understanding the paintings‘ somewhat anomalous existence exposes the mandalas as the

visual conflations of dharma and relic as body, and body as stūpa.

The mandalas not only privilege the text of the sūtras as the centerpiece of the paintings

and serve as visual commentaries on the nature of sūtra as dharma relic by referencing the

conventional partnering of relic and reliquary; they further layer the conflation of scriptural text

as the body of the Buddha through the structuring of the dharma in the form of a stūpa. In the

jeweled-stūpa mandalas, the complete imbrication of both materializations of the dharmakāya

serves as a visual treatise on the ultimate indivisibility of body, for one cannot behold the stūpa

without reading it as the sūtra, and it is equally as impossible to see the sūtra without regarding

the stūpa. The utter imbrication creates the visual capacity to constantly manifest and yet

195
subsume each of the dharmakāyas, the subject of chapter six‘s analysis of the text and image

relationships occurring within the mandala. This indivisibility portrays the conflation and

possible manifestations of the Buddha bodies. And more than that, the imbrication complicates

the conceptual boundaries of reliquary and relic, of exposed and hidden, of container and

contained; whereas reliquaries typically protected from damage and from sight the dharma relics

of the Buddha, such as the jeweled-stūpa sūtra canister which Seki Hideo 関秀夫 has suggested

once contained a sūtra scroll, 73 the jeweled-stūpa mandalas reveal an innovative perspective on

the functions of stūpa and sūtra. In the jeweled-stūpa mandalas, what was conventionally buried

inside and hidden from sight now builds and makes present the structure that historically housed

it.

Stūpa as a Site of Emanating Power

The bodily connotations of the stūpa certainly offer a compelling reason for its centrality in

mandalas rife with conflated notions of somaticity; however, the stūpa, as an active and highly-

charged monument in medieval Buddhism, functioned as a multi-dimensional icon. Therefore,

the central role of the stūpa in the mandalas reflects the varied symbolisms and functions of the

architectural reliquary. In this section, I suggest that the stūpa as a site of radiating salvific power

further warrants its prominent place in the mandalas.

I discuss stūpas as centers of salvific power manifesting the presence of the Buddha

where interaction between the cosmic divine and those who seek it is made possible. The

inherent potency, as bodies of the Buddha and as centers from which power radiated promising

proximity to the salvific presence of the Buddha, propelled stūpas as sacred spaces valued by the

73
Seki, ―Kyōzuka to sono ibutsu,‖ 23 fig. 30.

196
pious populace, clergy, and rulers. Again, a long tradition exists proclaiming the salvific grace of

the stūpa. As Schopen has shown in his study of the practice of burial ad sanctos, main central

stūpas marking important places in the life of Śākyamuni or containing corporeal relics were

often the objects of funerary anxieties and paradisiacal ambitions. 74 Around such monuments of

salvific potency, numerous smaller stūpas, many of which contain relics of their own—and as

Schopen points out, which were likely not those of Śākyamuni since relics of the Buddha were

most often inscriptionally indicated75—were constructed. There seems to be little spatial

planning or organization to the surrounding stūpas, and often newer stūpas were erected over

preexisting ones. As such, these examples of burial ad sanctos indicate an attempt to create a

cosmologically charged space for local funerary practices. Acting on both horizontal and vertical

axes, the smaller stūpas radiate out and down to create a multilayered three-dimensional space in

which people yearned for the proximity of the Buddha as manifested in the stūpa. Schopen notes

that the goal of this burial practice was to be near Śākyamuni, indicating a conviction that

funerary proximity to the Buddha as manifested in the stūpa gifted a soteriological impact.76 The

content of the dhāraṇī placed inside many of the smaller surrounding stūpas reiterated this

anxiety over death and karmic causation. Schopen offers a typical dhāraṇī found in this context:

―Moreover, if someone were to write this dhāraṇī in the name of another (who is deceased) and

were to deposit it in a stūpa and earnestly worship it, then the deceased, being freed (by that)

74
Schopen, ―Burial Ad Sanctos and the Physical Presence of the Buddha in Early Indian Buddhism.‖ A related
phenomenon in Japan is the practice of relic burials (Jpn. nōkotsu, Ch. nagu, 納骨) on Mt. Kōya around the grave of
Kūkai, except the space is horizontally arranged rather than the vertical piling of marker upon other marker. Reaping
the benefits of the funerary proximity does not require the whole of the body or all of the ashes of the deceased;
rather, the deposit can be mere hair or tooth. The promised bodily assumption of Kūkai at the advent of Miroku
encourages the popular practice of relic burials around Kūkai‘s grave, who resides in a state of samādhi there. Thus,
being buried next to the living presence of Kūkai ensured a beneficial rebirth. For more on this, see Nishiguchi,
Junko, ―Where the Bones Go: Death and Burial of Women of the Heian High Aristocracy,‖ transl. Mimi Hall
Yiengruksawan, in Engendering Faith: Women and Buddhism in Premodern Japan, ed. Barbara Ruch (Ann Arbor:
Center for Japanese Studies, University of Michigan, 2002), 428-35.
75
Schopen, ―Burial Ad Sanctos and the Physical Presence of the Buddha in Early Indian Buddhism,‖ 198.
76
Ibid., 201.

197
from his unfortunate destiny, would be reborn in heaven. Indeed, being reborn in the realm of the

Tuṣita gods, through the empowering of the Buddha he would (never again) fall into an

unfortunate destiny.‖77

The stūpa has played a critical and formative role throughout Buddhism. According to the

mystical origins of Shingon Buddhism, the Bodhisattva Vajrasattva (金剛薩埵 Jpn. Kongōsatta,

Ch. Jingangsaduo) was initiated into the esoteric teachings by none other than Mahāvairocana.

Vajrasattva then hid himself and the teachings away inside an iron stūpa where Nāgārjuna

discovered him and received the dharma transmission from within the protective walls of the

stūpa.78 This story was recorded in Kūkai‘s Record of Dharma Transmission, furthering its

popularity in medieval Japan.79 Another example illustrating the preservative and apotropaic

abilities of the stūpa comes from the story of the request by Zhiyi 智顗 (538-97, traditionally

considered the founder of Tiantai Buddhism) that a stone stūpa guard the texts he had written. 80

Imperial funerary practices certainly reflect the belief in the preservative powers of stūpas.

Numerous emperors and aristocrats requested the interment of their ashes within stūpas or for

stūpas to mark the place of their remains, as Haga Noboru‘s 芳賀登 study on the funerary

history of Japan reveals. 81 Even restoring a stūpa that has fallen into a state of neglect and

disrepair can deliver a favorable rebirth, as testified to in The Three Jewels (三寶絵 Sanbōe), a

text by Minamoto Tamenori 源為憲 (ca. 941-1011) for the princess, Sonshi Naishinnō 尊子内親

王 (ca. 966-985):

77
Ibid., 199. The dhāraṇī comes from the text, Raśmivimalaviśuddhaprabhādhāraṇī, in its Tibetan translation.
78
Stone, Original Enlightenment and the Transformation of Medieval Japanese Buddhism, 103. Relatedly, see
Okada Yukihiro 岡田行弘, ―Nagarunja to buttō suiha ナーガールジュナと仏塔崇拝,‖ Indogaku bukkyōgaku
kenkyū 印度学仏教学研究 48, no. 2 (2000): 1076-71 [reverse pagination].
79
Abé, The Weaving of Mantra, 221.
80
Stone, Original Enlightenment and the Transformation of Medieval Japanese Buddhism, 131.
81
Haga Noboru 芳賀登, Sōgi no rekishi 葬儀の歴史 (Tokyo: Yūzankaku 雄山閣, 1996).

198
 Long ago, there was an elder who lived after Vipaśyin Buddha has already entered
Nirvāṇa. He noticed a crack in the plaster at the back of a stūpa, so he prepared some
fresh plaster and repaired the crack, then he scattered sandalwood incense around the
stūpa, uttered a prayer, and went away. As a result, he did not fall into the Evil Realms
through ninety-one kalpas of rebirth but was always reborn as a Celestial or as a man
with fragrant body and a fragrant mouth…. Eventually, he became one of the Buddha‘s
disciples, and attained the state of an arhat. 82

Whether in doctrine, legend or religious practice, the stūpa has enjoyed a profound respect and a

revered sense of salvific power, cultivated over many centuries and across multiple borders. As

such it was the object of countless construction projects such as those discussed in the next

section.

Stūpa Commissions in Early Medieval Japan

Stūpa construction across medieval Buddhist societies enjoyed a long and sustained history. As

has already been discussed, numerous influential sūtras and commentarial texts encouraged such

building projects. Texts such as the Sūtra on the Merit of Building a Stūpa as Spoken by the

Buddha83 expound the meritorious potential of stūpa projects. The Buddha addresses Kannon

before the vast assembly of deities, saying:

Noble son, among the heavenly beings present here and all the living beings of future
generation, whoever is able to erect a stūpa wherever there is a place without one—
whether its form be exaltedly marvelous as to surpass the triloka or so extremely small as
an āmalaka fruit; whether its mast ascends to the brahma heaven or is as extremely small
as a needle; whether its parasol covers the great chiliocosm or is extremely small like a
jujube leaf.…84

And within this stūpa, if one enshrines a corporeal or dharma relic of the Buddha,

…this person‘s merit will be as great as the brahma heaven. At the end of his life, he will
be born in the brahmaloka. When his long life reaches its end in that realm, he will be

82
Edward Kamens, trans., The Three Jewels A Study and Translation of Minamoto Tamenori’s Sanboe (Ann Arbor:
University of Michigan, 1988), 269.
83
Jpn. Bussetsu zōtō enmyō kudokukyō; Ch. Foshuo zaota yanming gongde jing; 佛説造塔延命功德経; T. no. 699,
16: 801a11-b18.
84
Boucher, ―The Pratityasamutpadagatha and Its Role in the Medieval Cult of the Relics,‖ 9.

199
 born in the five pure abodes; there he will be no different than the gods. Noble son, of
such matters have I spoken—the magnitude of these stūpas and the cause of their merit.
You and all the heavenly beings should study and observe this. 85

In The Three Jewels, Tamenori recorded for the princess a short sūtra reporting the

dialogue—the Sūtra on Earning Merit for the Extension of Life through Stūpa Building86—

between King Prasenajit and the Buddha. 87 The King, convinced that in seven days time he will

succumb to death, beseeches the Buddha to extend his life. In this context, the Buddha preaches

the amazing benefits of stūpa construction: ―Among all excellent deeds, nothing exceeds the

excellence of stūpa building.‖88 He tells King Prasenajit of a young cowherd who was predicted

to die in seven days, but because the child built a stūpa of sand pebbles ―only as tall as the span

between his thumb and middle finger … his life was immediately lengthened by seven years.‖ 89

The Buddha also reveals to the King the enlightened message of a great sage who explained to a

group of mischievous children the exponential scale of the rewards of stūpa construction. The

sage tells them that building a stūpa of sand pebbles as tall as one hand will transform them into

an Iron Wheel-King in the next world; he goes on to link the size of the small stūpas to

exponentially increased rewards. 90 The Buddha grandly promises that ―[h]e who builds a stūpa

will be immune to poison for the rest of his life. His lifespan will be very long. He will not die of

unnatural causes; evil spirits will not come near him, and he will escape all his enemies and

assailants. He will never be ill, and his sins will be expunged.‖ 91

85
Ibid.
86
Jpn. Zōtō enmyō kudoku kyō; Ch. Zaota yanming gongde jing; 造塔延命功徳経; T. no. 1026, 19: 726a8-727c28.
87
Kamens, The Three Jewels, 279-81.
88
Ibid., 279.
89
Ibid.
90
Ibid., 280: ―When you build your stūpas of sand pebbles, make them as tall as your hand, and in the next world
you will become an Iron Wheel-King and rule one world. Make them two hands tall, and you will become a Copper
Wheel-King, and you will rule two worlds. Make them three hands high, and you will becomes a Silver-Wheel King,
and you will rule three worlds. Make them four hands tall, and you will become a Gold Wheel-King, and you will
rule four worlds.‖
91
Ibid.

200
 Tamenori joins the theme of small sand stūpas and their disproportionately large rewards

to the passage from the Lotus Sūtra which reads, ―There are even children who in play / Gather

sand and make it into buddha stūpas. / Persons like these / Have achieved the buddha path.‖ 92

This theme of children constructing stūpas for fun, yet unknowingly reaping great benefits, is

commonly represented in transformation tableaux and illustrated sūtra frontispieces, such as the

eleventh-century example of fascicle one of the Lotus Sūtra from Enryakuji illustrating children

building sand stūpas.93 The Lotus Sūtra at several points urges the construction of various types

of stūpas, usually to enshrine copies of the sūtra itself as a dharma relic. For example, the

Buddha instructs devotees, saying, ―O Medicine King! Wherever it may be preached, or read, or

recited, or written, or wherever place a roll of this scripture may occupy, in all those places one

is to erect a stūpa of the seven jewels, building it high and wide with impressive decoration.‖ 94

These proclamations and the many like them resulted in great building projects such as

those for the ailing princess in The Three Jewels. Tamenori records the activity of piling stones

into the shape of a stūpa, a group participatory practice common in the Heian period among lay

aristocrats occurring often in the spring. 95 These spring festivals frequently took place along a

river on the sixteenth day of the second month, the day after the anniversary of the Buddha‘s

parinirvāṇa.96 Tamenori, in a rare description of the festival, laments that ―there are ignorant

people who think of it merely as a pleasant excursion. They are responsible for setting the date of

the annual observance, and they are the arbiters of taste regarding the adornments upon the altars,

but in the evening they get drunk and collapse and tumble down the streets.‖97 But due to the

92
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 36. T. no. 262, 9: 8c24-25.
93
For an image, see Ariga Yoshitaka, ―Hokekyō-e 法華経絵,‖ Nihon no bijutsu 269 (1988): 64 fig. 97.
94
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 163. T. no. 262, 9: 31b26-28.
95
Kamens, The Three Jewels, 280 n2.
96
Ibid., 280 n.3.
97
Ibid., 279.

201
incredible merit derived from building stūpas, even those who view the occasion as an

opportunity to drink and carouse at night ―will reach the garden of merit, and they too shall plant

their own good roots there.‖98

The merit and salvific grace promised by stūpa construction and veneration propels

countless reliquarial projects in the early medieval period. The earliest recorded commission of

small stūpas of vast numbers occurred at Hosshōji on the twenty-third day of the fourth month in

1122 at the behest of Shirakawa.99 According to Hyakurenshō, a small hall was built to house the

300,000 stūpas measuring approximately three centimeters each. In 1125, Shirakawa made a

further commission of 10,000 small stūpas as a prayer for a peaceful childbirth for

Taikenmon‘in. 100 In a similar vein, Kujō Kanezane 九条兼実 (1149-1207) records in his diary,

Gyokuyō 玉葉, in the entry for the fourth month and twenty-fourth day of 1174 that Fujiwara

Motofusa 藤原基房 (1145-1231) commissioned 10,000 small stūpas to also pray for a safe

childbirth for his wife.101 And on the tenth day of the tenth month in 1178 Nakayama Tadamichi

中山忠親 (1131-95) recorded in his diary, Sankaiki 山槐記, that a court official ordered 15,000

small clay stūpas for his wife‘s safe childbirth.102 Small stūpa commissions were frequently a

part of a large project, such as the project by Toba and Taikenmon‘in in which 10,000 small

stūpas were produced for a series of ceremonies at Hosshōji, as recorded in Chūyūki.103

Much larger projects, such as the celebrated commission of 84,000 stūpas in emulation of

Aśoka‘s original dedication, were produced in great quantities. Tsuji Zennosuke claims that the

98
Ibid.
99
Tsuji, Nihon no bukkyō shi, 644. Kuroita, ―Hyakurenshō,‖ 52.
100
Tsuji, Nihon no bukkyō shi, 644.
101
Ibid., 645. Kujō Kanezane 九条兼実, Gyokuyō 玉葉, vol. 1 (Tokyo: Sumiya Shobō, 1966), 367.
102
Tsuji, Nihon no bukkyō shi, 645. Nakayama Tadamichi 中山忠親, Sankaiki 山槐記, in Zōhō shiryō taisei 増補史
料大成, vol. 27 (Tokyo: Rinsen Shoten, 1965), 146.
103
Fujiwara Munetada 藤原宗忠, ―Chūyūki 中右記,‖ in Zōhō shiryō taisei 増補史料大成, vol. 13 (Tokyo: Rinsen
Shoten, 1965), 293.

202
commissions of 84,000 small reliquaries became very fashionable during the Kamakura period,

functioning largely as memorials intended to offer repose for those killed as a result of the shift

in power to the Minamoto 源 clan, but also as prayers for good health and prevention of

disaster.104 For example, in 1197 Minamoto Yoritomo commissioned 84,000 deitōkyō (small clay

stūpas onto which a character from a sūtra is inscribed) to be constructed for the pacification of

those troubled spirits of the Hōgen disturbance 保元の乱 (July 28-August 16, 1156). Yoritomo

explains, ―we search here for the ancient tracks of Aśoka, constructing 84,000 jeweled-stūpas,

and believing [in the promise of] the benefits of wealth [i.e., wealth derived from the merit of

ritual], copy the Hōkyōin dhāraṇī [Skt. Karaṇḍamudrā dhāraṇī, 宝篋印陀羅尼] in all the sites

of spiritual power in the provinces [throughout the realm].‖ 105 Azuma kagami 吾妻鏡, a late

twelfth- through mid-thirteenth-century chronicle, documents several instances of projects of

84,000 reliquaries, many for the repose of the dead and prayers for health and prevention of

disaster. On the twenty-ninth day of the eighth month in 1203, the second Kamakura shogun,

Minamoto Yorie 源頼家 (1182-1204), dedicated 84,000 deitōkyō at Tsurugaoka Hachiman

Shrine 鶴岡八幡宮 in Kamakura for his recovery from illness. 106 Likewise, the fourth shogun,

Kujō Yoritsune 九条頼経 (1218-56), dedicated 84,000 deitōkyō for the prevention of disaster

and request for good health on the fourth day of the seventh month in 1241. 107 These are but a

sampling of the range and quantity of small stūpa projects of the early medieval period

manifesting the faith in the reliquaries to affect positive outcomes through their inherent salvific

and apotropaic power.

104
Tsuji, Nihon no bukkyō shi, 646.
105
Ruppert, Jewel in the Ashes, 237.
106
Tsuji, Nihon no bukkyō shi, 646.
107
Ibid.

203
 The stūpa as a body of the Buddha and worthy of worship and as a tremendous source of

salvific power (much like the motivation behind the prominence and privilege of dharma relics in

the mandalas) combined with the long tradition of stūpa construction and the culture of sūtra

transcription elucidates the significance and compelling factors behind the choice of the stūpa as

the dominant textual icon. As a monument dotting and adorning the medieval Buddhist sphere of

not only Japan, but many realms, the stūpa became a prolific emblem in visual culture which

perhaps found in the jeweled-stūpa mandalas its highest conceptual expression.

Conclusion: the Salvific Matrix of Text and Body

Jeweled-stūpa mandalas are visual nexuses of the somatic and textual matrix revealing the

imbrication of sūtra, dharma, body, relic, and stūpa in doctrine and praxis. Chapters four and five

worked to unpack the conceptual conflations of sūtra, relic, dharma, body, and stūpa as they

occur in doctrine and praxis, offering further evidence for the imbricated connections

establishing the mandalas. Throughout the visual and conceptual excavation of the paintings, I

have followed the salvific, textual, and somatic threads of the matrix as they emerged in each

section in an attempt to answer some of the fundamental questions regarding the inventive

privileging of scriptural text and the conceptual role of the stūpa as the dharma reliquary. In this

final section of the two tandem chapters, I pull together all the threads leading to the mandalas as

the nexus of this matrix and explain the conceptual fluidity manifesting and underpinning the

mandalas. In doing so I uncover the definitional imbrication that makes the mandalas so

exceptional.

Concerning the methodology involved in such an approach, I would like to highlight the

inherent tension between Japanese Buddhist doctrine, praxis, and culture, which so often treat all

204
of these elements as not only fluid and interconnected but at times as identical, and the demands

of scholarship and the process of linear explication in writing, which dictate that one analyze

each element distinctly and systematically. I hope I have resisted the urge to delimit these

interpenetrated concepts too strictly and have instead embraced and explored the ambiguities of

meaning and porous definitional boundaries, because it is this conceptual haze that produces the

visual manifesto on sūtra, body, and stūpa imbrication that is the mandalas.

Revealing the Salvific Strands

By analyzing the possible impetuses behind the manipulation of scriptural text into a textual

reliquary, I exposed a unifying strand revealing both sūtra and stūpa as repositories of significant

salvific power. What is perhaps a natural urge to understand and presentize the Buddha in his

seeming absentia creates a space to conceptualize areas of access. In the jeweled-stūpa mandalas,

sūtra and stūpa, as deeply comingling icons, are the access points, or more specifically, the

engenderers of the spheres of potent power.

As explored in chapter four, sūtra text embodies the essence of the Buddha, and therefore

has the ability to act on the world. At times scripture assumes human form to save the faithful in

need. At others, reciting, copying, and otherwise disseminating but a verse of scripture—even in

jest—is enough to envelop one in the salvific embrace of the sūtra. As testified to in doctrine,

setsuwa, commentarial literature, and in praxis, sūtra text encapsulated the Buddha-nature and

thus constituted a salvific sphere through which access to the Buddha and redemptive power was

possible. It is my contention that the desire to contact this repository of soteriological potency as

embodied in sūtra text encouraged the creation of the jeweled-stūpa mandalas.

205
 Stūpas, much like sūtras, have been celebrated as monuments of great salvific potential.

Worshiped with music, prayers, banners, and offerings of flowers, perfumes, and even land and

gardens, stūpas have enjoyed a long history of veneration. The Indian practice of burial ad

sanctos reflects the belief that stūpas possess great apotropaic and redemptive power. The urge to

inhabit in burial the space around the stūpa suggests that proximity to the stūpa affects a

favorable outcome, thus locating the stūpa as a site of radiating salvific power. This much is

attested to by the content of the dhāraṇīs found in the surrounding stūpas. The stūpa as a

protective and redemptive sphere is also at the root of the Japanese enshrinement of imperial and

aristocratic ashes within the reliquaries. And as demonstrated, building, guarding, and venerating

stūpas ensured protection and salvific grace for the devotee. Broadly, once we consider that

stūpas were the divinely chosen vehicle for the internment of arguably the most precious objects

in Buddhism—textual, corporeal, and contact relics—it is easy to see the potency believed to be

manifested in their architectural structure. As specially designated guardians entrusted with the

honor of relic enshrinement and their consequent ubiquitous presence dotting the religious

landscape of Japan, it is no wonder that the stūpa plays such a dominant visual role in the

mandalas and is a frequent icon in the larger visual culture of the eleventh through thirteenth

centuries.

Scriptural text and architectural reliquary embody transformative possibilities. Thus the

mandalas, as the nexus of this salvific web, visualize spaces of salvation. However, because of

the visual codependence of the sūtra and the stūpa, the mandalas manifest the two spheres as one

salvific space. In praxis, accessing this embodied power was made possible through the

transcription of sacred text and the construction of reliquaries (and their consequent veneration);

thus the paintings embrace both prescriptions, resulting in an innovative perspective on two of

206
the most common devotional practices. Following the strands connecting to salvation reveals the

mandalas as the nexus of a salvific matrix composed of both sūtra and stūpa.

Pulling the Textual Thread

Beyond written word and architectural form, the conflations within the dharma reliquary offer an

image exceeding the sum of its parts. The union of sūtra text and stūpa, with its references to

relics, dharma, and body invoked by this visual relationship, project a karmic confluence rather

anomalous to the mandalas. Putting aside for chapter six the complications of the conventional

text and image relationships inherent in the textualization of image, the mandalas offer multiple

perspectives on the definitional possibilities for the concept of sūtra text.

The very prominence of textuality in the mandalas dictates an investigation of the textual

thread running throughout the paintings. At the most cursory level, scripture is privileged as the

central and most dominant icon; however, this in itself is not all that remarkable. What sets the

mandalas apart from other text-based examples in visual culture is the innovative manipulation

of the sūtra into a textual reliquary. This privileging and prioritizing of sūtra expresses the

salvific power invested in sacred word, thus bringing together on the painting‘s surface the

salvific and textual strands of the web.

And because of the nondual conflation of scripture and dharma, sūtra text embodies the

ultimate truth that is dharma, and more specifically, the Buddha who is indivisible from the

teachings. Therefore, sūtra text is conceptually imbricated with the Buddha as dharma,

transforming word into potent relics. Following the reoccurring textual thread, we find that—

more than just written word—sūtra text is a dharma relic, and therefore, a manifestation of the

dharmakāya of the Buddha. As is evident in the pulling of these conceptual threads at work in

207
the mandalas, if one follows any one thread, inevitably they arrive at an intersection joining other

strands of the web. Indeed, I argue that the inextricably imbricated nature of the threads and the

visual manifestation of that conflation in the mandalas serve as the very basis of the paintings.

Surveying the Somatic Strand

Relating the theories of Buddha body and stūpa back to the mandalas and positioning the

paintings as the visual locus of the text, dharma, body, relic, and stūpa matrix reveals a diversity

of references to various manifestations of the bodies of the Buddha. The ambiguous and

multifaceted nature of body in the jeweled-stūpa mandalas is not unlike the controversy and

flexibility surrounding the buddhakāya doctrine in the literature. Therefore, precision of

conceptual definition is not the goal, for such a goal would inevitably be as illusory as it is

elusive. Rather, instead of wrestling with their ambiguity, the conflations of text, dharma, body,

relic, and stūpa in the literature and in the paintings should be seen as the unifying factor

impelling creation the jeweled-stūpa mandalas.

Pursuing the somatic strand within the mandalas reveals the ultimate relationship to body

shared by all conceptual components visualized in the paintings. Sūtra text as a manifestation of

the dharmakāya constructs the form of a stūpa, which is also an embodiment of the dharmakāya.

What is erected in the jeweled-stūpa mandalas is a textualized icon of vast salvific power and

presence: a dharma reliquary manifesting two visible somatic forms of the post-parinirvāṇa

Buddha. The dharma embodied in the scriptural text reveals another nuanced iteration of Buddha

body; and the rendering of sacred word as dharma relic, now constructing the architecturally-

conceived body of the Buddha, deconstructs to reveal body building body. With the emergence

of relics from the interior, hidden place within the reliquary to construct the monument that

208
previously secreted it away, the role of relics expands beyond its conventional boundaries and

urges viewers to contemplate the somatic potentialities when different bodily emanations build,

manifest, and yet subsume one another—when body erects another body.

Reflecting the Conceptual Conflation

Having traced the salvific, textual, and somatic threads of the mandalas, I suggest that this

reoccurring conceptual imbrication is rooted in the definitional fluidity of the concepts at work in

the paintings. The visual indivisibility of sūtra and stūpa, of relic and dharma, and ultimately of

the various emanations of body are all the graphic manifestations of the same nondual conflation

in doctrine. The very inextricability of the strands mirrors the transitive relationship experienced

in doctrine and praxis.

Perhaps an appropriate place to start is with the term dharma or dhamma. As has been

discussed in chapter four, dharma can be conceived of as the fundamental understanding of

reality and as the dharma that is taught; this creates a constantly interpenetrating notion of

dharma where dharma is the goal and the path, both the teacher and his message. Paul Harrison

concludes that this ambiguity is in fact built into the Pali term dhamma.108 Throughout the

Mahāyāna sūtras, dharmakāya is equated with other concepts such as śūnyatā (emptiness),

tathatā (suchness), dharmatā (reality of things), tathāgatagarbha (Buddha-nature), dharmadhātu

(truth concerning things, or realm of truth), and buddhadhātu (Buddha-principle), among several

others.109 Take for instance Nagao Gadjin‘s explanation of the equivalence of dharmatā and

dharmakāya as an example of the transitive relationship among the concepts:

108
Harrison, ―Is the Dharma-kāya the Real ‗Phantom Body‘ of the Buddha?,‖ 56.
109
For studies on the conceptual fluidity of these concepts, see William H. Grosnick, ―The Tathāgatagarbha Sūtra,‖
in Buddhism in Practice, ed. Donald S. Lopez (Princeton: Princeton University Press, 1995), 92-106; Habito, ―The
Notion of Dharmakāya,‖ 348-78; Harrison, ―Is the Dharma-kāya the Real ‗Phantom Body‘ of the Buddha?;‖

209
 [T]he word dharmatā (dharma-nature) came to be also used to represent the essence
itself of this dharma. Therefore, the dharma-kāya is the body of the dharma-nature as
well. Again, when the universe is conceived in the dimension of such dharma, the
universe is none other than the dharma-dhātu (dharma-realm). Being the true way of the
universe, the notion of dharma-dhātu is further identified with that of dharmatā or tathatā
(suchness) or even śūnyatā (emptiness).110

And as we saw earlier, many sources identify the perfection of wisdom (prajñāpāramitā) with

the Buddha and with śūnyatā, dharmatā, dharmadhātu, and tathatā to name a few. Within this

perfection of wisdom literature, the dharmakāya is characterized as the textual body of the

Buddha, and so sūtras are revealed to be the dharmakāya. As a further level of imbrication, the

stūpa is also identified with the dharma qualities possessed by a Buddha, and is thus equivalent

with the dharmakāya, which, we have seen, is synonymous with so many other concepts. The

terms śarīra, dhātu, and kāya can be all mean ‗body,‘ thus providing ample room for ambiguities

of meaning. This is not to suggest that these concepts are completely synonymous with one

another or to minimize the nuanced aspects of their meanings, but merely to point out the

inherent inseparability of one from the others.

In the jeweled-stūpa mandalas, these conceptual interpenetrations that are ultimately

inextricable from one another are rendered visible in the mandalas. The featured textual reliquary

very literally presents a body made of dharma with all of its conceptual layers. And within this

relationship, each seemingly independent manifestation can be deconstructed to reveal the

underlying identification with body, and thus, with the Buddha. Through this the mandalas

present a complex somatic reckoning, a visual treatise on the conceptual potentialities of these

Kajiyama, ―Stūpas, the Mother of Buddhas, and Dharma-body,‖ 9-16; Makransky, Buddhahood Embodied;
Makransky, "Controversy over Dharmakāya in India and Tibet;‖ Nagao, ―On the Theory of Buddha-body;‖ Suguro
Shinjō 勝呂信静, ―Hokekyō no budda-ron 法華経の仏陀論,‖ in Hokke Bukkyō no butsuda ron to shujō ron 法華仏
教の仏陀論と衆生論, ed. Watanabe Hōyō 渡辺宝陽 (Kyoto: Heirakuji Shoten 平楽寺書店, 1985), 61-110; and
Takasaki Jikidō 高崎直道, ―Dharmatā, Dharmadhātu, Dharmakāya and Buddhadhātu: Structure of the Ultimate
Value in Mahāyāna Buddhism,‖ Indogaku bukkyōgaku kenkyū 印度学仏教学研究 14 (1966): 919-903 [reverse
pagination].
110
Nagao, ―On the Theory of Buddha-body,‖ 27.

210
concepts. The graphic indivisibility captures the imbricated connections possible when

‗presentizing‘ the Buddha, for just as the understandings and manifestations of ‗Buddha‘ move

fluidly through a variety of concepts, these concepts also unite in him, forming intersections

where concepts merge and meanings synchronize. This definitional mutability discourages and

indeed prevents an exact, tidy, and fully delimited definition of these terms and their

relationships to one another; we cannot look at the dharma reliquary of the mandalas as just one

type of body, or as just a single manifestation of the Buddha‘s presence. Thus the mandalas

comprehensively interweave stūpa, dharma, relic, and sūtra—with all of their accompanying

conceptual luggage—to craft an innovative salvific matrix pulling together a variety of somatic

and textual threads.

211
 Chapter Six

Text and Image Issues in the Jeweled-Stūpa Mandalas

Introduction

Uttered aloud, committed to written word, or even inscribed within the mind, the nature and

many qualities of text have inspired volumes of philosophic discourse from Plato to

contemporary semiotic theorists. Clearly, the ubiquity of text across cultures and history has

made it a constant companion, yet the mutable borders of text confound strict definitions and

challenge interpretations which seek to limit its breadth. In this regard, text mirrors the diversity

and fluidity of Buddha body theory; perhaps it is ultimately the flexibility of text and body in

concert to enact a variety of functions and to assume a range of roles that makes possible the

jeweled-stūpa mandalas.

This chapter examines multiple issues that arise from the complex and imbricated

relationships of text and image in the jeweled-stūpa mandalas. I begin with an explicit

qualification of the use of semiotic perspectives in my examination of textualized images of early

medieval Japan. Although the dissertation has pursued the semiotic question concerning the

nature and function of text in the mandalas and in early medieval Japan, I have saved this direct

discussion until now because of semiotics‘ explicit role in the analysis of the functions of graphic

image and written word in the mandalas. I then present a survey of textualized images preceding

the mandalas in order to establish the visual community of the jeweled-stūpa mandalas but also

to accentuate their individuality. Following this summary I examine the ways of viewing

encouraged by the jeweled-stūpa mandalas. Necessary to this analysis is a consideration of the

212
roles of signifier and signified enacted in the paintings. This opens up a comparative space to

explore the role reversals experienced by text and image in the mandalas. Doing this leads to

questions about the nature of text and text‘s orality and materiality, along with a discussion of

the act of reading, particularly in the early medieval Japanese context as it relates to the

textualized reliquary and narrative vignettes.

Semiotic Perspectives

Semiotics provides an intriguing and enlightening framework through which to conduct analysis

of textualized images. Without knowing it or calling it by name, we often scrutinize and

understand visual culture through concepts and questions central to semiotic theory. Even a

project in which the basic components like signifier and signified are absent can still be revealed

to employ semiotic perspective in its analysis. Thus semiotics is not necessarily dependent upon

its self-conscious terminology, as it explores and tries to explain broad, perhaps even universal

questions about the world around us. Consequently, it is an interdisciplinary yet stealthy

approach.

The roots of semiotics can be traced back as far as perhaps any early thoughts on the

roles and limits of language. Curiously, semiotics continues to be a extra-departmental discipline

at most universities—cropping up usually within the established fields of art history, linguistics,

literature, film studies, anthropology, and philosophy. But through attempts to clarify its

objectives, questions, and fundamental assumptions by scholars such as Ferdinand Saussure

(1857-1913) and Charles Sanders Peirce (1839–1914), semiotics has emerged as a viable and

widely applicable methodology used across a variety of fields. This approach is at times

criticized as being unnecessarily convoluted, for complicating graspable concepts and

213
phenomena, and for ignoring the social, semantic, and syntactic context. Within Asian (art

historical) studies, the explicit use of semiotics has been limited. 1 Specifically, issues of

compatibility with Buddhism can be a concern. The transferability of theory crafted in Europe

and America to scholarship examining Asia is not an uncontroversial practice. However, if used

conscientiously semiotics can provide a useful theoretical framework for the analysis of Japanese

Buddhist visual culture.

At its root, semiotics seeks to decipher signs and the construction of meaning, defined

differently depending on the theorist. For example, the two theorists whose influence most

shaped semiotic discourse fundamentally disagree on the nature of the sign. Saussure

understands the sign as a binary relationship between the signifier and the signified, which for

him is not the actual object (physical though it may be) but rather the idea or concept of the

object, a limitation that denies any reality outside of the sign. 2 For Peirce, the sign is a

relationship involving the object or signified, the representamen or signifier, and how this

interaction is understood by the interpretant, an approach which accommodates specific space

for the role of interpretation by the person receiving the sign. 3 Setting this difference aside for a

more developed discussion later in the chapter, semiotics offers one method of decoding signs as

culturally embedded and situationally relevant occurrences and encourages the visual

interpretation of codes, an approach palatable to art historical examinations of textualized images.

In using semiotics to explore images like the jeweled-stūpa mandalas, I examined the

meanings of text in the mandalas and in their milieu, and in this chapter I continue to analyze

1
The clearest exception to this is Fabio Rambelli. For example, see Rambelli, ―Re-inscribing Mandala: Semiotic
Operations on a Word and Its Object,‖ Studies in Central and East Asian Religions 4 (1992): 1-24, ―The Semiotic
Articulation of Hosshin Seppō,‖ ―Texts, Talismans, and Jewels,‖ and Buddhist Materiality.
2
See for example, Ferdinand de Saussure, Course in General Linguistics, trans. Roy Harris (London: Duckworth,
1983), 66.
3
See for instance, Charles Sanders Peirce, Collected Papers (1931-58), ed. Charles Hartshorne and Paul Weiss, vol.
2 (Cambridge: Harvard University Press, 1932), 228.

214
how meanings are crafted by the often anonymous artists and viewers as well as by the sustained

and diverse relationship of image and culture in early medieval Japan. Broadly, this is an

examination into the construction of meaning and reality as portrayed in the jeweled-stūpa

mandalas. Moreover, semiotics grants the image an active and constitutive role in the creation of

reality because the image is more than what it represents, more than its material construction.

Because image is a representation, a particular perspective on the concepts it seeks to render and

not a one-to-one reproduction, it cannot advance a truly neutral transmission of reality but rather

can only contribute to the advancement of an advocated view of socially constructed meanings.

Indeed, many scholars such as Saussure, Roland Barthes, and Catherine Belsey maintain that

language, rather than reflecting reality, actually constructs it. As Belsey, a literary theorist, writes,

―If texts link concepts through a system of signs which signify by means of their relationship to

each other rather than to entities in the world, and if literature is a signifying practice, all it can

reflect is the order inscribed in particular discourses, not the nature of the world.‖ 4 Art historian

Ernst Gombrich makes a similar case in the context of art historical examination: ―…the

painter‘s starting-point can never be the observation and imitation of nature, that all art remains

what is called conceptual, a manipulation of a vocabulary, and that even the most naturalistic art

generally starts from what I call a schema that is modified and adjusted till it appears to match

the visible world.‖5 Our understandings of reality and the signs that establish our reality are

firmly embedded in our cultural context, a declaration which acknowledges the existence of

potentially innumerable realities.

Thus cultural lenses are required to decipher signs like the jeweled-stūpa mandalas as

much as is possible, for interpretations must not decontextualize the codes lest they become

4
Catherine Belsey, Critical Practice (London: Routledge, 2002[1980]), 43.
5
Ernst Gombrich, The Image and The Eye: Further Studies in the Psychology of Pictorial Representation (Oxford:
Phaidon, 1994), 70.

215
abstract, groundless images of beauty that remain sorely lacking in meaning. This scholarly

attitude is obviously already standard in contemporary art historical studies. Given the

inseparability of an image/sign from its cultural and historical context further necessitates an

intertextual examination. Intertextuality, while first coined by Julia Kristeva to describe the

interrelated nature of texts which refer in myriad ways to a multitude of other texts, has taken on

a life of its own and can be applied to studies beyond the textual, such as to visual images as in

this chapter. The idea of the interrelated, interdependent, and decidedly referential nature of texts

is captured exclusively neither in the term intertextuality nor only in the work of Kristeva. Other

scholars such as Roland Barthes, 6 Mikhail Bakhtin, 7 Claude Lévi-Strauss,8 and Michel Foucault

describe such interconnectivity. Foucault, in The Archaeology of Knowledge, presents the idea

succinctly: ―The frontiers of a book are never clear-cut: beyond the title, the first lines and the

last full stop, beyond its internal configuration and its autonomous form, it is caught up in a

system of references to other books, other texts, other sentences: it is a node within a

network…The book is not simply the object that one holds in one‘s hands; it cannot remain

within the little parallelepiped that contains it: its unity is variable and relative.‖ 9 Ryūichi Abé,

characterizing the Shingon understanding of signs writes, ―Things are never self-present, for they

have no ontological grounding, except for their infinitely regressive reference to other things in

their mutually referential network. That is, precisely because they are signs, things are of

dependent co-origination (Skt. pratītya-samutpāda; 縁起生 engishō [Ch. yuanqi sheng]), for

6
Roland Barthes, S/Z, trans. Richard Miller (Oxford: Blackwell, 1990).
7
Bakhtin‘s principle of dialogism requires a contextualized study of signs, which for him amounts to the literary
analysis of texts and their multiple meanings. Mikhail Bakhtin, The Dialogic Imagination: Four Essays, trans. Caryl
Emerson and Michael Holquist (Austin and London: University of Texas Press, 1981).
8
Claude Lévi-Strauss, The Savage Mind (London: Weidenfeld and Nicolson, 1962).
9
Michel Foucault, The Archaeology of Knowledge, trans. A. M. Sheridan Smith (London: Tavistock Publications,
1972), 23.

216
they are ‗empty‘ of essence and do not originate with any transcendental prime mover.‖ 10 While

this quote captures the referential quality and the cultural and historical location of signs, it also

reveals the emptiness of signs from some Buddhist perspectives. The principle of emptiness,

commonly held in most schools within Mahāyāna and expounded at length in the

Tathāgatagarbha Sūtras, denies phenomena an enduring essence and articulates the ultimate

emptiness of signs. In more extreme treatises on the nature of signs, semiotic structuralists reject

any fundamental connection between signifier and signified, maintaining an inherent, intractable

arbitrariness and emptiness of the signifier.

But as stated in the beginning of this section, a perfect correlation between semiotics and

Buddhism is not the goal, for such internal consistencies do not exist even within Buddhism or

the discourses of semiotics. This short section is merely intended to clarify some of the values

supporting the use of semiotics as a framework and its overall compatibility with an explication

of textualized Buddhist paintings. 11

Textualized Community

Before exploring the intriguing role reversal of text and image in the jeweled-stūpa mandalas, it

is necessary to survey earlier textualized images so as to give the mandalas a visual context and

to reveal the textualized community into which they were ushered, but also to highlight the more

complicated and imbricated relationships of text and image engendered in the mandalas. As is

pointed out in the book, Mojie to emoji no keifu 文字絵と絵文字の系譜, from booklets of waka

10
Abé, The Weaving of Mantra, 280.
11
W. T. J. Mitchell offers a call for increased attention to the role of language and word and image relationships in
art history inquiries: ―It must reflect on the relation of language to visual representation and make the problem of
‗word and image‘ a central feature of its self-understanding. Insofar as this problem involves borders between
‗textual‘ and ‗visual‘ disciplines, it ought to be a subject of investigation and analysis, collaboration and dialogue,
not defensive reflexes.‖ W.T.J. Mitchell, ―Word and Image,‖ in Critical Terms for Art History, ed. Robert S. Nelson
and Richard Shiff (Chicago: University of Chicago Press, 2003), 53.

217
superimposed upon scenes often of aristocratic life to elaborate inscriptions on Zen paintings,

image and text have enjoyed a longstanding intimate relationship as early as the eighth century

staged in diverse patterns on the visual plane of numerous examples of Japanese visual culture.

In these many textual expressions, the interaction of picture and word is a complicated and

variable affair, confirming that there is not just one type of text-image relationship in the visual

repertoire.12 The following paintings do not represent the whole of textualized images antecedent

to the mandalas, but merely establish a survey of the visual field out of which the mandalas

developed.

The countless explications and manifestations of sacred word in art, literature, and poetry

of early medieval Japan suggest scriptures to be open texts, capable of potentially endless re-

creation and reinterpretation. Indeed they necessitate constant and pious re-construction, as

claimed by Kūkai. Ryūichi Abé explains: ―Kūkai approaches the text as a yet-to-be bound—or,

perhaps more appropriately, never-to-be bound—constantly reworked manuscript. For Kūkai, the

text is not a book but a writing that remains open-ended.‖13 Similarly, literary theorist Terry

Eagleton has claimed that with each reading, a text is rewritten.14 The centrality of the textual

performance within the early medieval Japanese Buddhist ritual context is difficult to

overemphasize. In Karma of Words, William LaFleur describes the medieval epoch as a ―span of

time during which the literate people of the country held the classics of Buddhism to be the

ultimate norm—that is, the canon for integrating, interpreting, and judging a much wider range

of books and experiences they also accepted as valuable and, to a lesser degree, authoritative.‖ 15

12
Shibuya Kuritsu Shōto Bijutsukan 渋谷区立松濤美術館, ed., Mojie to emoji no keifu 文字絵と絵文字の系譜
(Tokyo: Shibuya Kuritsu Shōto Bijutsukan, 1996), 116.
13
Abé, The Weaving of Mantra, 276.
14
Terry Eagleton, Literary Theory: An Introduction (Oxford: Basil Blackwell, 1983), 12.
15
William R. LaFleur, The Karma of Words: Buddhism and the Literary Arts in Medieval Japan, (Berkeley and Los
Angeles: University of California Press, 1986), 11.

218
The early medieval period of Japan can be characterized as one penetrated by textuality. Beyond

hermeneutics and materiality, Buddhist texts were seen as embodiments of enlightenment.

According to Kūkai, ―[m]ountains are brushes, the ocean is ink / Heaven and the earth are the

box preserving the sūtras; / Each stroke of a character contains all things in the universe.‖ 16

Sacred word has been described as a ―microcosm‖ or ―holograph‖ of the Dharma Realm17—

through texts one accesses the Buddha.

Whether through the form of textual consumption known as kanjin-style interpretations

(―interpretations from the standpoint of the contemplation of the mind‖18) of sacred writings

popular in the Tendai school; the Shingon insistence on the ritualistic performance of both

esoteric and exoteric texts to unlock their meanings; the chanting of sūtra text or title widely

popularized by Amidist, Lotus, and Pure Land schools; the enshrining of sacred writings within

icons for ritualistic vivification; the practice of sūtra burials; or the pious transcription of

scripture;19 the performance, recreation, and enactment of sacred texts was woven into the fabric

of the religious and social context of early medieval Japan. Often textual performances resulted

in exquisite visual culture, as in the art of the sūtra presently examined. The many and inventive

artistic permutations of sacred word not only illustrate the concept of open texts, but embody and

manifest the great power of textual dharma. I believe the power of dharma relics—both salvific

and restorative—and the need to perform these texts through elaborate sūtra transcription

projects sparked their creation. In the following discussion, I explore the intertextuality of the

mandalas and the community of textual images of the eleventh through fourteenth centuries.

16
Rambelli, ―Texts, Talismans, and Jewels,‖ 73.
17
Rambelli, Buddhist Materiality, 119.
18
Stone, Original Enlightenment and the Transformation of Medieval Japanese Buddhism, 125.
19
This is not an exhaustive list, and the reader will likely be aware of further examples.

219
Divine and Profane Layers

The layers of sacred and mundane analyzed in this section render visual the belief in the potency

of textual dharma and the underlying, universal truth of nonduality. Examples such as the Lotus

Sūtra fans of Shitennōji, the Lotus Sūtra booklets, and the Golden Light Sūtra scroll visualize the

interpenetration of sacred writing with images of secular, aristocratic society. As Komatsu

Shigemi observes, the fans and booklets are visual testaments to the coupling of Heian period

aristocratic belief in the Lotus Sūtra and the pious expression of that faith. 20 In each example,

scenes of daily court life along with scenes from the world of commoners show through from

behind the superimposed textual dharma.

The deservedly famous twelfth-century Lotus Sūtra fans from Shitennōji are celebrated

as important testaments reflecting the intense belief in the Lotus Sūtra during the late Heian

period and as precious artifacts of decorated sūtras. It is possible that the original set may have

been donated by Fujiwara Yasuko 藤原泰子 (Kaya no In 高陽院), empress of Emperor Toba to

Shitennōji after her retreat to the temple for prayer in 1152. If so, this is a point of significance as

it would mean that the fans are the result of commission by a woman, thus transferring the merit

earned from the elaborate commission to her, and also that they preceded the famous Heike

nōkyō scrolls of 1164.21 The fans conform to the conventional rules for the copying of sūtra text

insofar as the scriptural text is structured into orderly lines of seventeen clear and intelligible

characters among twelve evenly spaced rows on each of the two sheets joined together by paste

in the center of the fan. The original set of ten volumes, standard for the transcription of the

Lotus Sūtra that include the opening and closing sūtras, are now dispersed into seven different

20
Komatsu Shigemi, ―Hokekyō sasshi ni tsuite 法華経冊子について,‖ Museum 81 (1957): 7.
21
Nara National Museum, Josei to bukkyō inori to hohomie 女性と仏教: いのりとほほみえ (Nara: Nara National
Museum, 2003), 232.

220
locations: Tokyo National Museum; Saikyōji 西教寺 in Shiga prefecture; Fujita Museum of Art

藤田美術館 in Osaka; Hōryūji; two private collections; and the largest amassment of the fans in

the collection of Shitennōji including volume one, six, and seven along with the bracketing

scriptures, the Innumerable Meanings Sūtra and the Sūtra of Meditation on the Bodhisattva

Universal Worthy.22

Rather than plain and lined or even brilliant blue or expensive, gold flecked paper, the

Lotus Sūtra fans, like those of Shitennōji, combine the graphic styles associated with illustrated

scrolls like Tale of Genji (源氏物語 Genji monogatari) with the recognizable structural and

kanji style of typical sūtra copies. The underdrawing 下絵 (shitae) combines images drawn by

hand and also rendered by woodblock print.23 On many of the fans appear figures dressed as

Heian period aristocratic court women known as the Jūrasetsunyo 十羅刹女 (Ten Demonic

Female Guardians). These defenders of those who maintain and honor the teachings of the

Buddha are strongly associated with the Lotus Sūtra.24 Most of the pictures describe life at the

imperial court and the lives of commoners at the close of the Heian period. The depictions are

varied, showing different seasons and landscapes, men, women, and children, the rich and the

poor, in an attempt render a broad scope of early medieval life. 25

Instead of segregating image from text, sacred word is layered upon such pictures as

aristocratic women at leisure, common life filling the streets, and the play of children. The

22
Tokyo National Museum 東京国立博物館, Kin to gin: kagayaki no nihonbijutsu 金と銀: かがやきの日本美術
(Tokyo: Tokyo National Museum, 1999), 274.
23
Nara National Museum, Josei to bukkyō, 232.
24
For more on the Jūrasetsunyo in Japanese art, see Nicole Fabricand-Person, ―Demonic Female Guardians of the
Faith: The Fugen Jūrasetsunyo Iconography in Japanese Buddhist Art,‖ in Engendering Faith: Women and
Buddhism in Premodern Japan, ed. Barbara Ruch (Ann Arbor: Center for Japanese Studies, University of Michigan,
2002), 343-82.
25
For much more on the histories of the fans and costumes appearing in them, see Akiyama Terukazu 秋山光和,
Yanagisawa Taka 柳澤孝, Suzuki Keizō 鈴木敬三, Senmen hokekyō no kenkyū 扇面法華経の硏究 (Tokyo:
Kajima Shuppankai, 1972).

221
layering of text upon image—for the pictures below would have been painted first—represent a

joining of two distinct media previously forced to inhabit different spatial realms of visual

culture, particularly evident in conventional illustrated handscroll sūtra transcriptions. This union

however does not correspond to any specific link between the two worlds, as it is noted that the

pictures and the sūtra‘s content are unconnected.26 But while text and image are combined into

one visual plane of the product—and this on its own represents an important marker in the

increasingly complicated and imbricating visual relationship of text and image—word and

picture still enact their own roles and maintain their functional and visual independence to a

large extent.

The Lotus Sūtra booklets offer the same treatment of text and image in a different yet

popular format, that of the bound booklet. On each page in which the sūtra is copied, seven lines

containing the standard seventeen-characters27 transcribed in ink present the fifth volume of the

scripture in the twelfth-century Lotus Sūtra booklet held in a private collection. Originally this

set was constructed as a detsuchōsō 粘葉装 (leaf book of pasted paper), but was later

reconstructed into a kochōsō 胡蝶装 (pasted leaf butterfly-style book).28 In total, there are forty

pairs, but among these sheets, there are passage omissions and sequential errors in the pages due

to errors in binding.29 Scenes of seasonal landscape and the private rooms of aristocratic

dwellings are populated by mostly court women carrying out daily routines. 30 The characteristic

techniques of the Heian period narrative works, such as fukinuki yatai 吹抜屋台 (―blown-off

roof technique,‖ a voyeuristic view into the private rooms of aristocratic life achieved through an

26
Tokyo National Museum, Kin to gin, 274.
27
For a discussion of the seventeen-character line in sūtra copies, see chapter two.
28
Kyoto National Museum, Koshakyō, 326. JAANUS notes that there is continued disagreement about the various
forms of bound books and their differences. See ―Kochōsō,‖ Japanese Architecture and Art Net Users System,
accessed December 1, 2010, http://www.aisf.or.jp/~jaanus/.
29
Kyoto National Museum, Koshakyō, 326.
30
See Nara National Museum, Josei to bukkyō, 231.

222
overhead, unobstructed vantage point), tsukuri-e 作り絵 (―built-up technique,‖ the procedural

construction of the scene in which the outlines are drawn in ink first, followed by the application

of color, and then the lines are redrawn and the faces applied), and the hikime kagihana 引目鉤

鼻 (―line for the eye, hook for the nose,‖ a restrained and abstract rendering of aristocratic faces

in which a diagonal line is drawn for the eyes, hook for the nose, and small, circular mouth), are

employed in this instance as they are in most of the other examples given in this section.

However, there does not appear to be any narrative connection between the scenes. The nature of

this set, largely supported by the prominence of women in the scenes, is thought to indicate that

it was for use by court ladies who could rest the small booklet in the palm of their hand. 31 The

script is very regular and legible, superimposed on these scenes of quiet interior life.

The Sūtra of Meditation on the Bodhisattva Universal Worthy booklet in the collection of

the Gotō Museum of Art provides another example of the intermingling of sūtra text with images

of earthly aristocratic scenes. This booklet conforms to the familiar arrangement of seven rows

of seventeen characters. The opening pages for volume five illustrate a picture of interior court

life at twilight, indicated by the halo of moonlight in the upper section of the image. A man

wearing his night robes in the lower right of the image is illuminated by the moonlight and the

faint light emanating from the fire he tends. Three court women populate the scene, one of whom

cradles a baby. Tiny white plum blossoms and the carpet of snow on the garden floor suggest a

chill in the air. The complicated patterns of the figures‘ robes further accentuate the displays of

wealth. Overall, the scene is highly atmospheric and dotted with seasonal and nightly references.

Interestingly, rather than sūtra text, the opening pages for both the third and sixth volume present

waka selected from the early tenth-century Kokin wakashū 古今和歌集 or the Anthology of

31
Kyoto National Museum, Koshakyō, 326.

223
Ancient and Modern Waka. It is thought that the sūtra was later copied as a memorial to the

owner of the uta-e 歌絵 booklet (the combination of waka with pictures that sometimes relate in

context).32

A vibrantly colored Lotus Sūtra illustrated scroll from 1266 within the collection of a

private individual once again merges the world of sacred text with images from the earthly

realm. 33 Superimposed lines of gold hold on average twelve, thick sūtra characters rendered in

black ink. However, at darker hued points in the underdrawing such as the brilliant blue water,

gold replaces the dark, black ink of the scriptural text; so while the two realms, pictorial and

textual, remain distinct in their overlapping, these brief switches directly acknowledge the

merger. On flecked, ornamental paper images of trees so red that the leaves almost read as

flames, flowing streams, and gently falling waterfalls establish the ambiance of an autumn day in

the mountainside. The scroll begins with elderly people and children and a woman beside a well.

A woman hunched over wearing a straw hat walks through the landscape. A small hermitage in

the mountains surrounded by a curving garden stream decorates the scroll. 34 An image of a nun

meditating before a seated Amida perhaps suggests that the underdrawing depicts a conversion

story: the entrance of a woman of intense belief into the clergy. 35 The scroll closes with the

image of a sotōba, and the copyist has inscribed the scroll with his wishes, signature, and date.36

The Menashikyō (literally, the ―eyeless sūtra‖) refers to an intriguing category of sūtra

scrolls firmly associated with Retired Emperor Goshirakawa wherein the text of the sūtra is

copied over a black ink, sketch-like underdrawing of pictures of interior court life and daily

32
Yamato Bunkakan 大和文華館, Nezame monogatari emaki bungei to bukkyō shinkō ga orinasu bi 寝覚物語絵
巻: 文芸と仏教織りなす美 (Nara: Yamato Bunkakan, 2001), 165.
33
For images of this scroll, see Nara National Museum, Josei to bukkyō, 95 plate 87.
34
Ibid., 231.
35
Ibid.
36
Ibid.

224
scenes, with the fascinating exception that most of the figures are left without eyes and noses.

The style of the pictures represents typical Heian period narrative illustrations; whether the

content of the underdrawing refers to a particular story is unknown. According to the colophon,

the Golden Light Sūtra version was transcribed on the first day of the fourth month in 1192. 37

Scrolls two, three, and four are the only remaining volumes. The Scripture that Transcends the

Principle version in the Dai Tōkyū Memorial Library in Tokyo, copied in 1192, offers a more

extensive albeit damaged colophon:

The Tonsured Emperor Goshirakawa and Nun [X]‘s painting, when not yet completed
[was interrupted by] the emperor‘s demise, whereupon the paper was used for copying
this sutra. The calligraphy [of the sutra text] is by [former] Major Counselor Master
Jōhen [and the] Sanskrit letters are by Master Jōken. In the eighth month of the fourth
year of the Kenkyū era [1192], this scroll was respectfully received from the abbot
[Shōken] by Shinken. 38

As Akiyama notes, Komatsu Shigemi suggests that the identity of the nun, sadly obscured by

damage to the scroll, could be Goshirakawa‘s consort, Takashina Eishi (d. 1216), the Lady of the

Tango Chamber.39 Akiyama also proposes that the Lady Kii could be the mystery woman due to

her strong connections with the monks associated with the scroll‘s production and ownership and

because she is often referred to as ―Kii the nun‖ in some documents. 40 What seems reasonably

secure is that Goshirakawa died before the completion of the picture scroll; and as a memorial

act intended to grant repose for the departed, the scroll was left unfinished and sūtra text and

Siddhaṃ-style letters were copied over the object closely related to the emperor, thus

establishing a karmic bond between the deceased and the redemptive powers of the sūtras.

37
Komatsu Shigemi, ―Menashikyō to sono shūhen 目なし経とその周辺,‖ Museum 60 (1956): 25.
38
Akiyama Terukazu, ―Women Painters at the Heian Court,‖ transl. Maribeth Graybill, in Flowering in the
Shadows: Women in the History of Chinese and Japanese Painting, ed. Marsha Weidner (University of Hawai‘i
Press, 1990), 167.
39
Komatsu, ―Menashikyō to sono shūhen,‖ 24-26.
40
Akiyama, ―Women Painters at the Heian Court,‖ 167-70.

225
 Does the imbrication of worldly image (with all its implications for sin and corruption)

with the potent, sanctifying word redeem the aristocracy, who would be the most likely audience

of these examples? Is the profane life then depicted in the background, representing the larger

illusory world of privileged society, purified and protected by salvific and apotropaic text? While

this is one way to interpret the combinatory textual images, applying the theory of nonduality

gives a more nuanced understanding.

Nonduality denies any ―ontological distinction between samsara and nirvana, or between

conventional and ultimate truth.‖41 Therefore, to quote LaFleur‘s discussion of Buddhist imagery

in poetry, even the symbols of our illusory world must ―be subjected to the…insistence that no

thing is ever merely a pointer or means for recognition of another thing.‖ 42 This is the

egalitarianism of signified and signifier, that a symbol is what it is and also what it represents.

The twelfth-century poet, Fujiwara Shunzei 藤原俊成 (1114-1204), in response to the dilemma

Buddhists faced in the composition of poetry due to the perceived impure qualities of verse,

counters with the argument that there can be no bifurcation of sacred and mundane. 43 The

interpenetration of one into the other, of reality and emptiness suggests that we cannot just

interpret the use of secular images as a mere juxtaposition or foil for the holiness of textual

dharma. In light of this, these paintings become instead the visual manifestation of the full

principle of nonduality. They encourage us to avoid the extremes of profane and holy, and reveal

instead the original enlightenment of all things, for ―this world is none other than the one of

tranquil light‖ (娑婆即寂光 shaba-soku jakkō).44

41
Stone, Original Enlightenment, 215.
42
LaFleur, The Karma of Words, 23.
43
Ibid., 91. Shunzei uses the Sūtra of Meditation on the Bodhisattva Universal Worthy as support for his argument.
44
This phrase is linked to Amidist thinkers. See Ibid., 97.

226
 Personal letters upon which are transcribed sacred scriptures often by the hand of a

grieving loved one, known as shōsokukyō 消息経, certainly embody the notions of potent

salvific word and nonduality. However, these letters also reveal a more intimate and private

appeal for salvation. The joining of these texts illustrate in a most literal way prayers for

deliverance, release from cyclical rebirth, and (re)merging with the dharma realm. 45 The first

recorded example of this practice, highly popular during the Heian through Muromachi periods,

is described in The True History of Three Reigns of Japan (三代実録 Sandai jitsuroku) in which

royal consort Fujiwara Tamiko 藤原多美子 (d. 886) copied the Lotus Sūtra on a letter written by

Fujiwara Seiwa 清和天皇 (850-80) upon his death.46 In its most basic structure, shōsokukyō fall

into two categories: the first is the copying of sūtra text onto the surface or reverse of a

departed‘s letter (this category can be broken into two separate types depending on whether the

sūtra is transcribed onto the front or back of the letter) and the second is the distillation of the

deceased‘s letters to make new paper onto which the sūtra is then copied. 47 Occassionally, hairs

of the deceased would be added during the distillation of the letter into sūtra paper to increase the

karmic connection.48

While sometimes ritualistically burned after production, many examples still remain

today. The thirteenth-century Kinji Amidakyō 金字阿弥陀経 of Rinnōji 輪王寺 in Tochigi offers

an elaborate shōsokukyō example.49 The paper is sprinkled with small gold and silver foil squares

(切箔 kirihaku), gold dust (砂子 sunago), and long, thin strips of gold foil resembling grass (野

45
Machida, ―Mukashi no shōsokukyō,‖ 32.
46
Kyoto National Museum, Koshakyō, 333. See also Tanaka, Nihon shakyō sokan, 28.
47
Yamato, Nezame monogatari emaki, 165.
48
Tanaka, Nihon shakyō sokan, 28.
49
For an image, see Tokyo National Museum, Kin to gin, 96 fig. 103.

227
毛 noge) and a flowering plant design are painted in gold and silver. 50 The sūtra text is also

copied in a bright gold between golden lines. The waka beneath is that of Fujiwara no Ariie 藤原

有家 (1155-1216), as recorded in the early thirteenth-century New Collection of Ancient and

Modern Waka (新古今和歌集 Shikokin waka shū).51 When compared to extant remains of

Ariie‘s handwriting, differences between authenticated examples and the waka here exist;

therefore, it is suggested that a person close to Ariie copied his waka and then transcribed the

Amida Sūtra over it.52 Great care is taken to preserve the integrity of the waka below. For

instance, the spacing of the lines organizing the scripture widen at places where the waka is

recorded so as to not overlap the delicate, black script. Additionally, space between the sūtra

characters is granted in order to allow the calligraphic writing below to emerge unobscured.

These amendments to the typical style of sūtra transcription reflect the context of this project and

symbolize the respect for the remains left behind that come to embody the presence of the

departed. As Machida Seishi points out, retaining the calligraphy of the deceased increases the

commemoration.53

The early fourteenth-century Lotus Sūtra scroll copied by Emperor Fushimi 伏見天皇

(1265-1317) in a kuyō or memorial ritual for his father, Emperor Gofukakusa 後深草天皇

(1243-1304), is transcribed on a letter from Emperor Gofukakusa, either for the forty-ninth day

death anniversary or to mark the passage of one year since the death.54 In this example, instead

of copying the scripture directly atop the letter, the reverse of the letter is used. In this way, both

texts coexist in the same realm, yet remain autonomous. This preservation of both the hand of

50
Ibid., 272.
51
Ibid.
52
Ibid.
53
Machida, ―Mukashi no shōsokukyō,‖ 32.
54
Kyoto National Museum, Koshakyō, 333 and 245 fig. 138 for an image.

228
Fushimi and Gofukakusa allow the two scripts to be compared. Fushimi, a celebrated

calligrapher, copied the Amida Sūtra in an elegant, semicursive style in great contrast to the

heavy, cursive script of Gofukakusa that demonstrates a slanting axis towards the upper right. 55

Another example meriting brief mention is the Mahāvairocana Sūtra56 copied in the

thirteenth century atop a letter of highly cursive and flowing script.57 The Mahāvairocana Sūtra

is a rather unusual choice of scripture, because most shōsokukyō use either the Lotus Sūtra or the

Pure Land sūtras (浄土経 Jōdokyō) such as the Amida Sūtra.58 While neither the sūtra copyist

nor the author of the letter is known, it is clear that great effort was taken in the production of

this shōsokukyō. The margins of the original letter were cut, the paper gently beaten, and mica

sprinkled before the scripture was copied. The thin, calligraphic lines of the original letter show

through behind the rather thick and regular characters of the sūtra. Another shōsokukyō

remarkable for its choice of scripture is the twelfth-century Bussetsu tennyo jōbutsukyō 仏説転

女成仏経 of the Tokyo National Museum. 59 Interestingly, the small, golden characters of the

sūtra and black, flowing kana of the letter exist in a visual equilibrium, neither overemphasized

nor dominant. This scripture, along with the Lotus Sūtra‘s twelfth-chapter, the ―Devadatta

chapter,‖60 preaches the entrance of women into nirvāna.61 Importantly, both were used at the

end of the Heian period in memorial services for women according to the entry on the nineteenth

day of the eighth month in 1077 within Suisaki 水左記, the diary of Minamoto Toshifusa 源俊房

55
Ibid., 333.
56
Jpn. Dainichi kyō; Ch. Dari jing; 大日経; T. no. 848, 18: 1a4-55a4.
57
For an image, see Kyoto National Museum, Koshakyō, 244 fig. 137.
58
Ibid., 333.
59
Yamato, Nezame monogatari emaki, 76 fig. 44.
60
Jpn. Daibadatta bon; Ch. Tipodaduo pin; 提婆達多品.
61
Yamato, Nezame monogatari emaki, 165.

229
(1035-1121).62 It is likely then that this shōsokukyō was produced for the memorial services of an

aristocratic woman.

The writing of sūtras, letters, and waka upon used paper is a widely used technique, not

always associated with the practice of shōsokukyō. Often times, sūtra copies were appropriated

for temple business and letters of diverse purposes were sent out on the back of scripture because

paper was a precious commodity. For instance, in a letter to the monk Shinkai 審海 of Shōmyōji

称名寺, Ninshō 忍性 appeals to Shinkai and his fellow practitioners to participate in a ritual at

Gokurakuji 極楽寺 for a visiting monk named, Ashōbō 阿性房 of Kyoto‘s Kanshūji 観修寺.63

In other cases, sūtra copies were used as paper for the writing of unconnected waka, such as the

poem of Fujiwara Kintō 藤原公任 (966-1041), one of the thirty immortal poets, copied onto the

Lotus Sūtra.64 Interestingly, letters were occasionally written on the reverse side of the Engishiki

延喜式, the rules of the ritsuryō state, as exemplified in a letter by Minamoto Kaneyuki 65 and

another letter thought to be brushed by a woman. 66 Other examples joining text upon text clearly

have no intended religious connection at all, thus demonstrating that this practice had a wide

application. 67

Shōsokukyō are not the typical objects of art historical study, probably because they are

purely textual compositions. However, I have included several examples of shōsokukyō for two

62
Ibid.
63
Nara National Museum, Kamakura bukkyō kōsō to sono bijutsu 鎌倉仏教: 高僧とその美術 (Nara: Nara
National Museum, 1993), 189-90, and fig. 34 for an image. Similar examples are the Kōzanji monjo 高山時文書,
see NHK Promotion プロモーション, ed., Yoshitsune ten: Genji, Heishi, Ōshū Fujiwara-shi no shihō 義経展: 源
氏, 平氏, 奥州藤原氏の至宝 (Tokyo: NHK Promotion, 2005), 60 fig. 59; and the Shinkyō shojō 真教書状, see
Nara National Museum, Kamakura bukkyō, fig. 95.
64
Yamato, Nezame monogatari emaki, 75 fig. 43.
65
Tokyo National Museum, Sho no shiho: nihon to chūgoku 書の至宝: 日本と中国 (Tokyo National Museum,
2006), 201 fig. 98.
66
Ibid., 186 fig. 86. Similarly, see the Daigozōjiki 醍醐雑事記 in Tokyo National Museum, Kokuhō daigoji ten 国
宝醍醐寺展 (Tokyo National Museum, 2001), entry 67.
67
For an example of this, Akihagijō 秋萩帖, see Tokyo National Museum, Sho no shiho, 184-85 fig. 85.

230
reasons. Firstly, they directly correspond to a larger theme of the project: the role, power, and

ritualistic functions of sūtra text during the ninth through thirteenth centuries. As compositions

revealing the anxiety of death and the hopeful prayers of the departed‘s loved ones, shōsokukyō

are a testament to the perceived power of scriptural word to redeem and save after death.

Secondly, these combinatory texts offer a fascinating opportunity to analyze the visual roles of

the texts, challenge the standard understandings of text and image, and discuss the issue of

foreground versus background and the implications of such dynamics on the definition of textual

images. The visual relationships at work in the shōsokukyō manifest an enlightening perspective:

sacred text is seen through the prism of the secular world and the secular text appears inside the

realm of the pure and sacred. Within this second reason I explore three different visual scenarios

occurring within these combinatory texts.

The first scenario occurs where text is layered upon text, yet the copyist of the sūtra takes

care to avoid eclipsing the original script below. Thus the personal letter or waka remains intact

and often is still capable of fulfilling its original and intended purpose. Visually, private text and

sacred text coexist in an equitable balance, neither subsumed nor privileged, but sharing a visual

plane where both texts are legible and unhindered: a purely textual image that communicates an

abiding respect for the original text and its creator.

The second happens when the notions of background and foreground are challenged in

the shōsokukyō. Examples where the scripture is transcribed onto the reverse side of the secular

text subvert the dynamic of primary and dominant versus secondary and subordinate. In an

object where the primary, intended focus is unclear, a fluid and flexible viewing situation is

created. This is particularly true when the quality of the paper or the heaviness of the brush

causes a simultaneous viewing of both texts, one through the other. In such a circumstance, the

231
vague yet constant presence of one within the other manifests the visual interpenetration of

secular and sacred. Notions of background and foreground are also challenged in examples of the

first type mentioned above. When the shōsokukyō equalizes the presence of both texts through

compositional structure, thereby jettisoning the mechanisms of prominence, the restrictions

imposed by the categories of background and foreground are eluded.

The third relationship is the utter breakdown of the typical textual components of the

letter or waka. In examples of this type, either through the processes of preparing the paper for

the sūtra transcription such as by beating the paper, by trimming the original structure, and other

altering procedures or by utterly obscuring the initial text with that of the scripture, the secular

text is stripped of its original textual functions—namely, the communication of a message or the

reading of verses. In the context of the shōsokukyō, the original functions transition from

standard textual ones to those of the religious context: to bear witness to the departed in

memorial rituals. The text, still present but rendered illegible, assumes the role of image within

this textual composition. In instances where the personal writing of the deceased is completely

dissolved and the paper refashioned to make a new surface for the copying of the sūtra, the

original text is stripped down to its fundamental, material basis.

A related practice is that of the Genji kuyō 源氏供養 or memorial service for the Tale of

Genji in which a scroll containing the text is washed clean and the Lotus Sūtra is copied atop the

recycled paper, reflecting the belief that writing and reading secular texts was a sin. Therefore, in

an attempt to right the transgressions of both the readers and the author of secular texts, the

purifying Lotus Sūtra is used. In analyzing this form of ritualistic sūtra copying as well as the

practice of copying scripture onto funeral clothes called kyōkatabira 経帷子, Fabio Rambelli

writes, ―as the profane substance of one‘s body (blood) becomes the support (and the signifier)

232
of the scriptures, so the material substance of Tale of Genji (the paper out of which the book is

made) becomes the support of the Lotus Sūtra.‖68

Imbricated Images

In this next section examining the textualized community that produced the jeweled-stūpa

mandalas, I look at the practice of ashide 葦手 (―reed-hand script‖) within sūtra frontispieces.

This type of disguised script was often found in marsh-like landscapes where kanji and kana

formed simple images such as rocks, reeds, coast lines, and birds in flight; Komatsu Shigemi

provides a rich analysis of the motifs assumed by ashide in his study of the Heike nōkyō.69 In this

study, he finds that certain kana are routinely chosen to construct particular and specific pictures

because of their inherent shape lends them naturally to certain common forms. For example, ka

か and na な often form rocks and sa さ and fu ふ usually form flying birds. 70 The practice of

ashide extended broadly into many different formats and contexts of writing, such as that of uta-

e or poem-pictures. While the script crafted by the ashide often could be constructed into

meaningful passages of sūtra text or waka, ashide also had a purely decorative function as well.

Because of the breadth of ashide in visual culture of the ninth through thirteenth centuries

(indeed the practice continued much later), I primarily focus on ashide in the context of sūtra

transcription and select a couple of examples to highlight the textualized character of the images.

In 1164, Taira Kiyomori commissioned one of the most elaborate and sumptuous of the

ippon kechienkyō 一品結縁経, a lavish type of sūtra transcription in which a single scroll is

dedicated to one chapter. The Heike nōkyō is a set of thirty-three scrolls consisting of twenty-

68
Rambelli, Buddhist Materiality, 250.
69
Komatsu, Heike nōkyō no kenkyū, 819-29.
70
For many more such examples, see Ibid., 823-29.

233
eight rolls of the Lotus Sūtra as well as single rolls of the Amida Sūtra, Heart Sūtra, Innumerable

Meanings Sūtra, and Sūtra of Meditation on the Bodhisattva Universal Worthy. Kiyomori

composed the petition scroll in his own hand and invited thirty-two members of his family and

important retainers to craft one each. The set was then dedicated to Itsukushima Shine. A

particularly interesting example of text and image relationship comes from the frontispiece of

chapter, ―Former Affairs of the Bodhisattva Medicine King‖ frontispiece. 71 The frontispiece

depicts an Amida Welcoming Descent in the upper left corner, rays of divine light issuing forth

from his ūrṇā (a tuft of hairs between his eyes). Next to Amida floats a small lotus throne,

supported by a wispy, purple cloud. An aristocratic female figure leans on an armrest,

appropriately reading the twenty-third chapter of the Lotus Sūtra.72 The sumptuously decorated

scene is flecked with gold and silver details.

While the text of the sūtra still remains separate from the frontispiece, thus maintaining

the standard segregation, portions of the image itself are constructed from highly calligraphic

forms of kanji and kana. The disguised text would challenge aristocratic viewers to locate the

obscure message hidden among the images. As Julia Meech-Pekarik points out, the sūtra open

before the woman reads, ―The woman who hears this sutra and keeps this chapter of the Previous

Life of the Medicine-King Bodhisattva will not be a woman in her next life. After my extinction

…‖73 The passage quoted from the chapter ends there but is continued among the rocks and reeds

of the textualized image. For example, above the woman‘s head is the katakana for moshi モシ

(if), below her knees the katakana for no ノ of kono コノ (this), beside her right knee the

71
For an image, see Egami, ―Sōshokukyō,‖ 29 fig. 31.
72
This chapter is a particularly important chapter for women because it promises salvation to all women who hear
the sūtra and uphold its teachings.
73
Julia Meech-Pekarik, ―Disguised Scripts and Hidden Poems in an Illustrated Heian Sutra: Ashide and Uta-e in the
Heike Nōgyō,‖ Archives of Asian Art 31 (1977): 52-78. My discussion of the frontispiece borrows heavily from
Meech‘s article.

234
katakana characters arite アリて (there is), floating toward Amida the kanji for kokoni myōjū 此

命終 (when this life is over), further along framing the shoreline the kanji for sunawachi 即

(instantly), below the Buddha‘s right knee and hidden among the lotus petals the kanji-hiragana

phrase anraku sekai 安楽せかい (world of happiness), and the kanji for umaru 生 (to be born)

very clearly rests atop the pedestal.74 With some sleuthing the erudite viewer discovers the

masquerading text and the remainder of the phrase may be completed. Transforming the

graphically expressed woman and Amida Buddha into text, the phrase is finally finished, albeit

in shorthand form: ―The woman who hears this sutra and acts according to the teachings of it …

will [immediately] be able to be reborn, after her life in this world ... on the jeweled seat in the

lotus flower blooming in the World of Happiness where Amida Buddha lives surrounded by

great Bodhisattvas.‖75 The metamorphosis from pictorial form into the text of the sūtra

transforms the woman‘s body; she becomes part of the sūtra but also is the instrument for the

writing of the sūtra which in turn ―rewrites her body as that of a Buddha.‖ 76

More than just the dissemination of dharma, the ―Former Affairs of the Bodhisattva

Medicine King‖ frontispiece presents the doubly enforced message of salvation through not only

graphic image, but through an image composed of hidden textual meaning. Much like the

discussion of original enlightenment and nonduality above, we find here the visualization of the

world as text, a revelation that the scripture penetrates all manner of things in our world. Thus it

is appropriate that the Heike nōkyō offers a glimpse of the deep penetration of Buddhism into

cultural pastimes like these literary games requiring considerable literacy prowess.

74
Ibid., 74.
75
Ibid.
76
Eubanks, ―Rendering the Body Buddhist,‖ 332.

235
 One last example deserving mention is the Collection of Japanese and Chinese Verses

[with ashide] (葦手和漢朗詠抄 Ashide wakan rōei shū) copied by Fujiwara no Koreyuki 藤原

伊行 (d. 1175) in 1160 according to the colophon attached to the second scroll. 77 The disguised

script hides among the natural features common to waterside landscapes. Ka か, na な, no の,

and other kana form reeds, rocks, and other motifs. 78 In this case too, the ashide forms riddles

associated with key words or phrases from the verses listed atop the simple drawings. The visual

relationship of text and image in examples like the Heike nōkyō and Fujiwara Koreyuki‘s Wakan

rōei shū parallel the dynamic witnessed in the jeweled-stūpa mandalas, with the important

exception of scale. Text in both cases forms the pictorial image; however, because of the hidden

nature of the text in the ashide examples and because the graphic quality of the ashide is

emphasized over the textual, I differentiate the treatment of word in these textualized images by

classifying the ashide as pictorialized text. And while I categorize the jeweled-stūpa mandalas

within this same category of text and image relationships, ―imbricated images,‖ I separate the

discussion of the mandalas into its own section to follow.

Empowered Inscriptions

Before returning to the jeweled-stūpas mandalas, it is necessary to briefly introduce another

development in the text and image relationships of early medieval Buddhist painting: that of the

privileged text. Here I wish to emphasize the utter abandonment of graphic picture and the

assumption of strictly textualized compositions where word now paints the picture that graphic

image once captured. To highlight this phenomenon, I examine the Great Mandala of Nichiren

77
Tokyo National Museum, Sho no shiho, 340 and 218-20 fig. 108 for images of the scroll.
78
Ibid., 340.

236
Shōnin 日蓮上人 (1222-1282).79 In the Great Mandala, text through calligraphic expression

becomes the image.80

Both celebrated and reviled, Nichiren was a fervent—some even say rabid—proponent of

the Lotus Sūtra as the supreme Buddhist authority within which all other doctrines and praxis are

subsumed.81 Nichiren‘s advocacy of the Lotus Sūtra as the ultimate authority and the sūtra‘s

emphasis on text and language-oriented practice82 is reflected in his promotion of the sūtra‘s

daimoku (title) as the mantra, namu myōhō rengekyō (homage to the Lotus Sūtra). He famously

wrote, ―It is better to be a leper who chants Namu-myōhō-renge-kyō than be chief abbot of the

Tendai school.‖83 According to Nichiren, the title of the scripture contained within its five

characters the power to realize buddhahood in this very body (即身成仏 Jpn. sokushin jōbutsu,

Ch. jishen chengfo),84 much like ―All Buddhas of the three time periods and ten directions

invariably attain Buddhahood with the seed of the five characters myōhōrengekyō.‖85 Thus, the

daimoku served all dimensions of religious practice and expression and should be the follower‘s

constant practice.

The Great Mandala grew out of Nichiren‘s advocacy of the Lotus Sūtra as the supreme

authority, reflected in his daimoku practice. In an essay written in 1260, Nichiren responds to a

question about the appropriate object of worship for those who are dedicated to the Lotus Sūtra:

―First of all, as to the object of worship, you may use the eight rolls of the Lotus Sūtra, or a

79
For more information on Nichiren, see Stone, Original Enlightenment, 239-356.
80
For good images of Nichiren‘s Great Mandalas, see Kyoto National Museum, Nichiren to hokke no meihō
hanahiraku Kyōto machishū bunka 日蓮と法華の名宝: 華ひらく京都町衆文化 (Kyoto: Kyoto National Museum
and Nihon Keizai Shinbunsha, 2009), 54-58 figs. 43-47 and 74-76 figs. 65-67.
81
Stone, Original Enlightenment, 261.
82
Stone, ―Not Mere Written Words,‖ 160.
83
Stone, Original Enlightenment, 254.
84
Ibid., 241.
85
Ibid., 271.

237
single roll, or one chapter, or you may inscribe the title and make it the object of worship.‖ 86

This passage reflects the germinating seed for the Great Mandala.

As such, Nichiren‘s mandala depicts the venerated title of the scripture in calligraphic

script running vertically down the center of the scroll. The names of Śākyamuni and

Prabhūtaratna as well as the names of other deities populating the ten realms flank the central

daimoku, recreating the calligraphic assembly at Vulture Peak.87 Nichiren individually inscribed

the mandalas for his disciples, instructing them to practice the invocational daimoku before the

Great Mandala because through this unity of contemplation and invocation the practitioner could

enter the enlightened space of the mandala. 88 Indeed, the 1280 Great Mandala stored in

Myōhonji and copied by Nichiren is said to have hung beside his deathbed,89 perhaps providing

comfort or focus during the last few hours of his life.

Nichiren‘s mandala represents yet another twist in the relationship of text and image.

Graphic image, in conventionalized form, is completely abandoned in the Great Mandala. We

find no anthropomorphic Buddha figures, no text restructured to create an image. Instead

Nichiren fashions a calligraphic inscription, itself an image of exceptional fluidity and grace.

What emerges after brush has left paper is not just written word, but a portrait of the infinite

soteriological powers of the Lotus Sūtra, in effect a textual image. The Great Mandala surveyed

in this section manifests a different, more textualized dynamic between word and picture. Rather

than the cohabitation of text and image, the Great Mandala demonstrates a complete usurpation

of picture by text in a realm traditionally dominated by graphic image. Other examples of

86
Jacqueline Stone, ―Chanting the August Title of the Lotus Sutra: Daimoku Practices in Classical and Medieval
Japan,‖ in Re-Visioning "Kamakura" Buddhism, ed. Richard Karl Payne (Honolulu: University of Hawai'i Press,
1998), 152.
87
Stone, Original Enlightenment, 260.
88
Stone, ―Chanting the August Title of the Lotus Sutra,‖ 152-53. See also Takeda Nichikatsu 竹田日濶, ―Nichiren
no mandara to kihonzon tono kankei ni tsukete 日蓮の曼荼羅と其本尊との関係に就て,‖ Indogaku bukkyōgaku
kenkyū 6, no. 1 (1958): 152-53.
89
Stone, Original Enlightenment, 281.

238
empowered inscriptions in which text is privileged occur with increasing frequency in the

Muromachi through Edo periods, as discussed in the concluding chapter.

Role Reversals

Perhaps in hindsight it is possible to identify a trajectory from the conventional illustrated sūtra

scrolls clearly separating scriptural word from the graphic frontispiece, through the textualized

images popular in the Heian period, to the utterly imbricated jeweled-stūpa mandalas. Certainly,

it is possible to see an increasingly innovative approach to the role of text in the realm of

painting. More and more, text creeps into the domain of the image—so often a purely visual

space—blanketing pictures of aristocratic leisure and daily activities; discreetly crafting rock,

reed, or marshy shoreline; and finally becoming image itself, as in the jeweled-stūpa mandalas.

In this section, I examine the ways in which the conventional roles of text and image are

challenged, and indeed, reversed.

What appears from afar as inert, simple line constructing the central stūpa deconstructs

upon closer examination, betraying the solidity and continuity of what was first perceived to be

architectural line. The disaggregation of the shape revealed to be textual characters from the

scriptures occurs in a couple of steps, exposing the inherent structural dynamism of the mandala.

First, an overall transformation occurs during the initial stages of viewing in which the static line

dissolves into tiny, individualized characters forming the body of the stūpa, establishing that this

central icon is in fact a textual reliquary erected of dharma. Upon more intimate inspection, the

dynamic arrangement and twisting movements of the characters emerge as the eye attempts to

trace a line of text, stumbling upon characters that twist and turn and dangle from roof eaves. It is

at this point that the stūpa relinquishes much of its pictorial quality and becomes instead lines of

239
twisting text, character stacked upon character: an emergent text. Thus with close scrutiny, the

image of the stūpa dissolves into text, and with distance, again reemerges as picture in an

oscillating and fluid transformation.

The text of the sūtra, due to the incredibly small size of the characters and its structural

manipulation into a graphic image, jettisons its potentially expository role. While the text

continues in order, moving from top to bottom and right to left, a reading of the scripture for

content becomes infeasible. No longer for exegetical analysis, text instead becomes an artistic

device and an emblem of redemptive and soteriological power. That sacred scripture was not

always intended to be consumed character by character in each of its visual manifestations

testifies to the diverse functions and values of sacred text. In this regard, the mandalas

correspond to a wider set of occupations and purposes embodied in texts of early medieval

Buddhist Japan. As explained in chapter two‘s discussion on the possible functions of the

jeweled-stūpa mandalas, the sūtra text of many transcription projects was likely not meant to be

read or chanted. Rather, the purpose of the practice was the act of copying itself. Therefore, texts

like the jeweled-stūpa mandala‘s stūpa, elaborate sūtra scrolls like those of the Heike nōkyō, and

scriptures copied for the purposes of burial assume roles beyond the borders of exegetical

reading; and, as introduced in chapter one, texts in the early medieval Buddhist context often

exceed the limits of hermeneutics. Flipping through a sacred text, albeit ritualistically, granted

the participant great merit. Textual encounters, even fleeting or frivolous ones, 90 had the ability

to convey tremendous apotropaic and salvific merit as well as the more earthly ambitions

associated with the authoritarian and social value of the texts. However, because an actual

reading or even perusal of the sūtra is made impossible by the small size and gentle acrobatics of

the characters, the mandalas manifest a further transformation of text: the intensification of the
90
Examples of this sort were discussed in chapter four‘s examination of setsuwa.

240
visual properties of word. Thus, the scripture of the dharma reliquary experiences a reversal of

the conventional roles of text transcending that of typical sūtra copies: the textual stūpa becomes

graphic image in function and appearance.

Likewise, the narrative vignettes surrounding the text-as-image stūpa undergo a role

reversal as well. Because the sūtra text relinquishes its discursive properties, the vignettes

assume the role of content transmitter through the graphic manifestations of the sūtra‘s didactic

episodes. The arrangements of the illustrations seem to obey no discernable order. Often the

narrative episodes depicting a particular tale are not even grouped together. Similarly, the

narrative sequence does not correspond to the order of the episodes as they occur in the sūtra.

And the location of the illustrations in the narrative space does not relate to the section of text

beside which it is painted. Occasionally, a cosmological order is followed with paradisiacal

scenes grouped toward the top of the mandala and depictions of hell toward the bottom of the

scroll. Despite the seemingly random, even chaotic narrative assembly, the familiarity of the

illustrations and the text from which they derive allow viewers to read the graphic manifestations.

Image, imbued with textuality, can be examined and read for doctrinal insights.

As an example of such narrative vignette reading, I take a few episodes from the twenty-

third chapter of the Lotus Sūtra, ―Former Affairs of the Bodhisattva Medicine King,‖ as depicted

in the seventh scroll of the Ryūhonji set and explain the way in which the narratives are read for

their doctrinal content.91 The chapter begins by describing the extraordinary appreciation of and

devotion to the Lotus Sūtra felt by Medicine King Bodhisattva for the understanding and

enlightenment gained through hearing the recitation of the scripture. He offers gifts such as

flowers, oils, and sweet scents to the Buddha Pure and Bright Excellence of Sun and Moon who

preaches the Lotus Sūtra from paradise. However, Medicine King Bodhisattva is unsatisfied with
91
For images of these scenes, see Miya, Kinji hōtō mandara, 194-97.

241
these offerings: ―After he had made this offering, he arose from samādhi and thought to himself,

‗Though by resort to supernatural power I have made an offering to the buddha, it is not as if I

had made an offering of my own body.‘‖92 In order to communicate his extreme piety and

gratitude towards the Lotus Sūtra, he commits self-immolation and his body burns for a period of

twenty thousand years. 93 Due to his great piousness, the Buddha reconstitutes Medicine King

Bodhisattva. He immediately returns to the presence of the Buddha, bowing in obeisance and

offering prayers. The Buddha informs Medicine King Bodhisattva of his decision that same night

to enter parinirvāṇa, the physical death of the body and the passage into nirvāṇa. The scene of

parinirvāṇa is found in the lower left corner of the mandara; this episode illustrates the Buddha

lying prone on a raised dais, surrounded and worshiped by disciples of the Buddha, heavenly

deities, and mythical animals. The vignette above the parinirvāṇa scene is a depiction of the

Buddha‘s instructions to Medicine King Bodhisattva to build reliquaries for his relics which are

then to be housed and disseminated in 84,000 stūpas: ―‗After my passage into extinction,

whatever śarīra there may be I entrust to you also. You are to spread them about and broadly

arrange for offerings to them. You are to erect several thousand stūpas.‘‖94 Following the

pictorial illustrations in a clockwise path, the next episode describes the creation of Buddha

relics: the cremation of the Buddha on the funeral pyre. It is from this act that corporeal relics

were formed. Directly above this scene is a series of pictorial similes, representing the promised

gifts and great benefits of the Lotus Sūtra described in this chapter. Along the right side of the

mandala and in the middle of the long, narrow band of pictorial illustrations are located two

92
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 270. T. no. 262, 9: 53b4-5.
93
Encouragement for one to commit autocremation can also be found in Chinese texts, such as the Fanwang jing
(The Brahma Net Sūtra): ―In accordance with the dharma he should explain to them all the ascetic practices, such as
setting fire to the body, setting fire to the arm, or setting fire to the finger.‖ James A. Benn, ―Where Text Meets
Flesh: Burning the Body as an Apocryphal Practice in Chinese Buddhism,‖ History of Religions 37, no. 4 (1998):
299. Also, see James A. Benn, Burning for the Buddha: Self-Immolation in Chinese Buddhism (Honolulu:
University of Hawai‘i Press, 2007). For more on self-immolation in Japan, see Tsuji, Nihon no bukkyō shi, 636-39.
94
Hurvitz, Scripture of the Lotus Blossom of the Fine Dharma, 272. T. no. 262, 9: 53c14-5.

242
more episodes detailing the past life of Medicine King Bodhisattva. After completing his task,

Medicine King Bodhisattva offers his forearms because he remains unsatisfied by his donations of

the reliquaries and stūpas. The two episodes along the right of the architectural structure illustrate

this moment of the Bodhisattva‘s fervent gift; Medicine King Bodhisattva extends his forearms

engulfed in flames toward the three stūpas as the passionate gift of his body. 95 Below this scene,

Medicine King Bodhisattva, seated in the lotus position, is depicted moments after his offering has

been made, for slender wisps of smoke trail from his arms. Worshippers gather round his figure. This

is the last scene illustrated from the twenty-third chapter of the Lotus Sūtra. This role assumed by

image is not unlike the function performed by other narrative descriptions found in illuminated

frontispieces or transformation tableaux. But in a context where text is included, it is unusual that

the entire task falls to the responsibility of visual depiction alone.

Thus in order to study the many parables and episodes within the scripture, the viewer is

compelled to confront the Lotus Sūtra tales, not through discursive examination, but visually, by

interpreting the narratives—in effect by reading the pictures. In this way, image in the form of

the pictorial vignettes assumes a textual role. And it is when the combined visual effects of the

boundary pushing mandalas are considered that we realize the full consequence of the role

reversals occurring and reoccurring in a single painting and the rarity of this sort of combinatory

composition.

In the jeweled-stūpa mandalas, text—no longer functioning for exegetical analysis but

instead assuming the role of an image—is manipulated into the form of a stūpa, evoking

questions concerning the conflations of reliquary, dharma relics, and Buddha body as addressed

in the previous two chapters. Conversely, image in the form of the narrative vignettes are imbued

95
For more information on the gift of the body, see Reiko Ohnuma, ―The Gift of the Body and the Gift of the
Dharma,‖ History of Religions 37 (1998): 323-59, and Head, Eyes, Flesh, and Blood.

243
with textuality, becoming the repository of doctrinal insights through which the stories of the

sūtras are read.96 Thus text and image experience a role-reversal of their conventional functions.

As Mimi Yiengpruksawan asserts, ―doctrine and image at once reinforce and subvert one another,

and … the friction so generated enriches readings of all Buddhist objects be they words or

pictures.‖97 As such, it is possible to read the role reversal evinced in the jeweled-stūpa mandalas

as a subversion of text by image and vice versa. This section worked to expose the imbricated

roles of two previously distinct media; text forms pictures and image reads as text, creating a

vacillating, surreptitious relationship between written word and pictorial image.

Mojie

The central icon blending picture and text belongs to a particular category of images called mojie

文字絵. According to the broadest definition in the dictionary, Kōjien 広辞苑, mojie are pictures

written of text, in essence textual pictures. 98 The catalogue produced in conjunction with an

exhibition on such paintings, Mojie to emoji no keifu, seeks to alter and expand the definition of

mojie and the other related category of pictures, emoji 絵文字 (pictorialized text).99 Specifically,

mojie is characterized by two types of occurrences. The first type consists of tiny characters

arranged to create a larger design. 100 The second category of mojie includes the integration of the

characteristics of a letter or kanji, such as its shape, into part of a picture; 101 this includes hidden

script, such as ashide. Clearly, the jeweled-stūpa mandalas fall into the first group of mojie. To

96
I return to the issue of reading later in the chapter.
97
Mimi Hall Yiengpruksawan, ―Illuminating the Illuminator: Notes on a Votive Transcription of the Supreme
Scripture of Golden Light (Konkōmyō saishō ōkyō),‖ Versus 83/84 (2000): 116.
98
Shinmura Izuru 新村出, ed., Kōjien 広辞苑, 4th ed. (Tokyo: Iwanami Shoten, 1991), 2537. The dictionary divides
mojie into three categories: 1) Pictures drawn of text; 2) The shape of a warrior and other such figures drawn using
text to which pictures of a head, arms, and legs are added; 3) Ashide.
99
Shibuya Kuritsu Shōto Bijutsukan, Mojie to emoji no keifu, 119-20.
100
Ibid., 120.
101
Ibid.

244
my knowledge, the textual reliquary of the jeweled-stūpa mandalas were the first example of

mojie in Japan. However, this particular, even peculiar, visual format is actually a broadly

occurring artistic phenomenon. For example, evidence surviving from the ninth century shows

that masorah scribes in Israel created Hebrew micrography. 102 Masorah is a system of marginal

biblical notes which documents each word in the Hebrew Bible, recording how many times it

appeared and where.103 It gradually developed into an artistic form, taking the shape of birds and

other animals as well as people and biblical figures. Several scholars offer explanations for the

artistic development of masorah.104 Claude Gandelman posits that since Jewish law forbids

drawing the human body, text functions in Jewish micrography as a form of subversion. 105 By

using text to construct the anthropomorphic form, the artist has circumvented the law by writing

the human form and is therefore ―theologically safe.‖106 Such may be the case with an illustrated

sheet from the Song of Songs where text composes the anthropomorphic form of King Solomon

in an illustration from a book on circumcision. 107 Similar to Buddhism which holds dharma as

relic, some branches of Judaism perceive the written word to be mystically animated. Stanley

Ferber writes that, ―A major aspect of this German branch of Jewish mystic thought was the

endowment of magical properties to the word, letter, their various combinations, and their

enumeration.‖108

102
Leila Avrin defines a microgram as an image whose outlines are made of text, whereas a calligram is an image
composed entirely of text. See Leila Avrin, ―Hebrew Micrography: One Thousand Years of Art in Script,‖ Visible
Language 18 no.1 (1984): 87.
103
Ibid., 90.
104
See for example the works by Stanley Ferber, Leila Avrin, and Claude Gandelman.
105
Claude Gandelman, ―By Way of Introduction: Inscriptions as Subversion,‖ Visible Language 23, no. 2/3 (1989):
161.
106
Ibid.
107
Leila Avrin, Micrography as Art (Jerusalem: The Israel Museum, Jerusalem, 1981), 56. This ink on paper image
is currently in the Staats- und Universitätsbibliothek, Hamburg and is dated to 1819-20. For an image, see Ibid.,
plate 91.
108
Stanley Ferber, ―Micrography: A Jewish Art Form,‖ Journal of Jewish Art 3/4 (1977): 20.

245
 Other visual parallels can be found in Sufi art. The calligrams of Amadou Bamba, the

saint of the early twentieth century around which Senegalese Mouridism developed, exhibit

relationships of text and image similar to the jeweled-stūpa mandalas.109 Mourides, a designation

for members of the Senegalese Sufi movement, champion the sole photograph of Bamba—

hiding his face in shadow and intensifying the esoteric and saintly nature of Bamba—translating

his image into calligrams. Described as ―‗self-consciously esoteric‘‖ and ―‗the inner dimension

of Islam,‘‖ for the purposes of this short section, I adopt the definition of Sufism as ―a situated

knowledge and localized practice, for its paths lead to Paradise through the teachings of

particular saints who lived in particular places at particular times.‖ 110 Bamba is said to have been

astonishingly prolific in his lifetime. One story tells of a person ―who inadvertently entered the

Saint‘s chambers to find Bamba‘s ten fingers transformed into quill pens, all writing at once.‖111

Mourides consider a poem written by Bamba a passport to paradise: if someone dies with a copy

of the poem on them, God will permit entry into heaven. 112

Icons of Bamba are believed to ―actively bless, heal and protect people.‖113 Such images

served as talismans. These calligrams use holy words to capture the portrait of the saint. Often

passages from the Quran cover the face of Bamba, expressing his sainthood and dissolution into

Allah. Or his own poems are used, increasing the efficacy of the image, for his poetry is

considered miraculous and redemptive. Such portraits build his body with his own words,

fashioning a body of poetry. The blessing, ―There is no God but God, and Muhammad is the

109
One example is a reverse-glass calligram of Amadou Bamba by Serigne Gueye, after a photograph sold on the
street, and composed with glass, pigment, cardboard, and tape in 1993, now in the UCLA Fowler Museum of
Cultural History. For an image, see Allen F. Roberts and Mary Nooter Roberts, A Saint in the City: Sufi Arts of
Urban Senegal (Los Angeles: UCLA Fowler Museum of Cultural History, 2003), fig. 1.20. Another interesting
example is the 1913 photograph of Bamba rendered in calligraphy by an unknown artist on ink and color on paper in
1999, now housed in a private collection. For an image, see Ibid., fig. 7.5.
110
Ibid., 22.
111
Ibid., 168.
112
Ibid., 24.
113
Ibid., 18.

246
Messenger of God,‖114 constructs the face of Bamba in one image. Across the eyes and cheek of

the saint are written the names of Allah and Muhammad, ―suggesting his effacement into the

Word of God and his proximity to God and to the Prophet.‖115

Two works by Xu Bing 徐冰 (b. 1955) offer one final instance from contemporary China.

Born to a librarian and a professor, Xu Bing grew up with access to the restricted shelves of

Beijing University library. This early interest in the written word has remained an integral part of

Xu Bing‘s work as a print maker.116 Landscript plays with conventional notions of landscape

painting and calligraphy, building the landscape using characters such as tree and mountain in

repetition, thereby writing the landscape and merging the cultural practices of poetry, calligraphy,

painting, and seal into one.117 In the words of Xu Bing, ―Since they originally come from the

same root, I am merely uniting them again.‖ 118 In this drawing, text becomes image, returning

characters to their original function as pictographs.

The issues raised by the close collusion of these two media, merged to create a new

textual image that is neither strictly word nor picture, spark interesting perspectives on the role of

both. Textual pictures also provoke questions concerning the inherent gulf between words and

images. Is it possible that words can express meanings that elude capture in graphic images?

W.J.T. Mitchell characterizes the relationship of the word and image as two countries that share

a long history of relations but speak different languages.119 Gombrich declares that ―statements

114
Ibid., 59.
115
Ibid.
116
Xu Bing, during a talk given at the University of Kansas, April 26, 2007.
117
For an image, see Jerome Silbergeld and Dora C.Y. Ching, eds., Persistence/Transformation: Text as Image in
the Art of Xu Bing (Princeton: P.Y. and Kinmay W. Tang Center for East Asian Art in Association with Princeton
University Press, 2006), 108 fig. 13.
118
Xu Bing, ―An Artist‘s View,‖ In Persistence/Transformation: Text as Image in the Art of Xu Bing, ed. Jerome
Silbergeld and Dora C.Y. Ching (Princeton: P.Y. and Kinmay W. Tang Center for East Asian Art in Association
with Princeton University Press, 2006), 109.
119
―The domains of word and image are like two countries that speak different languages but have a long history of
mutual migration, cultural exchange, and other forms of intercourse.‖ Mitchell, ―Word and Image,‖ 53.

247
cannot be translated into images‖ 120 and that ―pictures cannot assert.‖121 According to Michel

Foucault, there exists an untraversable chasm eternally separating word and image. He believes

written word and graphic image run parallel to one another, that what is expressed in text cannot

be given visual form while retaining the original meaning of the text. The same fractured

communication exists when visual form is described by word. The chasm prevents full

expression of one by the other.122 However, Foucault finds hope in calligrams, believing that

they bring ―a text and a shape as close together as possible‖ 123 by simultaneously invoking and

conflating both avenues of communication: written discourse and visual representation. Foucault

writes, ―Pursuing its quarry by two paths, the calligram sets the most perfect trap. By its double

function, it guarantees capture, as neither discourse alone nor a pure drawing could do.‖124 If we

apply this Foucauldian rubric to the analysis of the jeweled-stūpa mandalas, the paintings

become a perfect union of Buddhist expression, combining and unifying both sūtra text and

visual reliquary and illustrating—textually and pictorially—the body of the Buddha. Using the

framework provided by Foucault, the conflations theologically and visually amplify and augment

the significance and potency of the mandala.

Rather than merely reinforcing or confirming what is already known, visual culture has

the potential to reveal new perspectives and creations. The jeweled-stūpa mandalas certainly

expand not only visual but conceptual possibilities. The few examples discussed above offer a

glimpse of the widespread practice of constructing textual images across time, place, and culture.

The role reversals witnessed in the mandalas provide new ways to think about the limits and

120
Gombrich, The Image and The Eye, 138.
121
Ibid., 175. For more on the limitations of pictures, see the chapter, ―The Visual Image: Its Place in
Communication,‖ Ibid.
122
Michel Foucault, The Order of Things: An Archaeology of the Human Sciences (New York: Vintage, 1973), 9.
123
Michel Foucault, This Is Not a Pipe, transl. and ed. James Harkness (Berkeley; Los Angeles; London: University
of California Press, 1983), 20-21.
124
Ibid., 22.

248
potentialities of text and image in a visual and Buddhist context. It is because of the union of

word and picture in the dharma reliquary that the proposed matrix from chapters four and five is

possible; the conflation of sūtra text and the visual image of the stūpa manifests a complex

treatise on body and salvation accessible because of imbrication of text and image. In the next

section, I explore the issues involved in viewing the mandalas and the concepts of signifiers and

signifieds at work in the textual reliquary.

Ways of Viewing

Viewing intricate paintings such as the jeweled-stūpa mandalas is a complex task. As described

earlier in this chapter, the recognition process resolving the complicated and imbricated elements

of the painting occurs in a series of steps. Visual cognition is a cumulative and continuous

development. The intertextuality of the mandalas with earlier and contemporary paintings

discussed here and between the sets themselves creates a referential system of acquired,

sustained, and emergent understandings about how objects should look and what they mean. The

ways in which we view paintings is a cognitive practice requiring cultural and historical

awareness at best, but also some optical, cognitive processes that operate perhaps without our

recognition of the various steps that occur. Of course, critical to issues of viewing visual culture

and the intertextuality of objects and the ideas that underpin them is the examination of the

historical and cultural context, the topic of chapters four and five in which I proposed a salvific

matrix theory. This section proposes to analyze the visual consequence of the paintings. Because

of the visual complexity and interdependency of text and image, the jeweled-stūpa mandalas

provide a fascinating opportunity to explore the ways in which we approach and read paintings.

249
 The culture of viewing examines how one approaches or is made to view a painting. 125

With paintings of such elaborate and interconnected word and image forms, the audience must

negotiate their viewing experience. To quote Claude Gandelman, ―Inscriptions can also be said

to represent the ‗performative‘ aspect of the work of art in the literal meaning of this word; that

is they are used to direct the gaze of the observer to specific spots within painting and are part of

the manipulative strategy of the painter.‖126 Working from the theories of J. L. Austin, 127

Gandleman describes a form of kinetic subversion, meaning that the inscriptions cause a

perlocutionary effect which forces the viewer to perform some action or confront the paintings in

a prescribed way. 128 He believes this ―sort of viewing produced by the intrusion of the ‗semiotic

enclave‘ is what one might call the ‗syncopated‘ viewing of a picture; that is, it causes a

syncopated vision and a constant interchange and exchange of vantage points.‖ 129

Jeweled-stūpa mandalas oblige such syncopated viewing. The audience, from a distance,

may not register the central stūpa as architecture constructed of written dharma, but upon closer

inspection, the imbrication of image and text forces the viewer to both see and read the textual

reliquary. Because of its visually disparate parts,130 approaching the mandala requires syncopated

viewing and demands a give and take of vantage points. Seeing the textual stūpa from afar gives

little indication that it is in fact composed almost entirely of written word, a recognition which

only comes from close inspection. However, in reading closely the textual characters, the shape

of the stūpa dissolves. Given the quantity of illustrations flanking the central icon, the narrative

125
Gandelman, ―By Way of Introduction,‖ 140-169.
126
Ibid., 140.
127
J. L. Austin proposes the concepts of locutionary act, illocutionary act, and perlocutionary act. J. L. Austin, How
to Do Things with Words (Oxford: Oxford University Press, 1976). Austin‘s ideas have made a profound impact on
studies of language and image. For instance, Lopez incorporates some of the concepts from Austin‘s scholarship
within a Buddhist study.
128
Gandelman, ―By Way of Introduction,‖ 146.
129
Ibid., 148.
130
That is to say, the mandala is constructed of disparate visual parts: the graphic image of the stūpa built of written
word and the combination of pictorial narratives and sūtra text.

250
vignettes also require significant optical attention. Seeing and reading the mandala as a whole

becomes impossible in this light. According to Mieke Bal, the viewing of every painting creates

a new event;131 and, the performance—optical and cognitive—required by the mandalas offers a

rare viewing experience.

It is the complex text and image relationships at work in the jeweled-stūpa mandalas that

produce this particularly complicated set of viewing possibilities. Fundamental to any

consideration of the issues of viewing and interpreting such an imbricated textual image is an

examination of the notions of signifiers and signifieds. However, whereas much of the discussion

concerning these two staple components of semiotic analysis focuses on the space or gap

between the two; Derrida in his discussion of difference explains, ―By definition, difference is

never in itself a sensible plenitude. Therefore, its necessity contradicts the allegation of a

naturally phonetic essence of language. It contests by the same token the professed natural

dependence of the graphic signifier.‖ 132 The composition of the mandalas problematizes this

dynamic. Inherent in the standard assumptions of signifier and signified is that only partial

signification is ever possible. But within the context of Buddhist imagery, this limitation is not

necessarily present. It is the very combinatory action taken in the mandalas which creates an

imbricated image that allows for the various forms of the Buddha body to manifest. And whereas

the presence of the signifier typically marks the absence of the signified, in the jeweled-stūpa

mandalas, dharma relics assume the roles of both signifier and signified and the body constructed

of this sacred text is that of the Buddha. In consequence, what are being signified are different

understandings of body, the possibilities of language, in essence the salvific matrix, manifested

131
Mieke Bal, Reading “Rembrandt” Beyond the Word-Image Opposition (Cambridge: Cambridge University
Press, 1991).
132
Jacques Derrida, Of Grammatology, trans. Gayatri Chakravorty Spivak (Baltimore and London: The Johns
Hopkins University Press, 1976), 53.

251
through dharma relics. Therefore in these paintings, the constant slippage of signifier into

signified and reference into referent escapes the rigid duality imposed by semiotics.

Jacqueline Stone writes, ―poetry, even art itself, is not a second-level representation of a

higher, ‗religious‘ truth but, when approached with the proper attitude, is equivalent to Buddhist

practice and is the expression of enlightenment.‖133 From at least the ninth century onward in

Japan, Buddhist thought in general encouraged the nondual view of the phenomenal (事 ji) with

the principle (理 ri), in Tendai referred to as ‗phenomena are none other than the true aspect‘ (現

象即実相 genshō soku jissō). This relationship between actuality and the representation of that

truth—and even all things in this illusory world—is manifested in the visual culture again and

again. William LaFleur pronounces that ―the use of symbols suggests that language is two-tiered.

When transformed into a symbol a thing remains what it was and becomes something else as

well.‖134 The mandalas, therefore, are not simply symbols of the body of the Buddha and salvific

word, but actively partake and manifest the presence of that which they embody. Clearly, the

materiality of the text within the jeweled-stūpa mandalas is fundamental to its functions and

potentialities. In the next section, I continue the examination of the visual impact and material

significance expressed in the mandalas.

Materiality

Certainly, texts were valued beyond their discursive function for their performative qualities and

for their material manifestation of the ‗immaterial,‘ the physical expression of which constituted

various systems of value, from economic to symbolic and religious currency. 135 Indeed the

133
Stone, Original Enlightenment, 45.
134
LaFleur, The Karma of Words, 17.
135
Rambelli, ―Texts, Talismans, and Jewels,‖ 52-53.

252
hermeneutical sense of reading was not the primary purpose of sacred texts, for the vast and

influential meanings of word extend far beyond what was directly signified.136 Texts should not

be reductively understood only through their hermeneutic or discursive properties because this

ignores the many dimensions of their lives, materiality, orality, and performativity. The various

interpretations and innovative uses of Buddhist texts reflect their polysemic nature. Barthes

characterizes the interpretation of texts in what he describes as the Nietzschean sense of the term,

claiming that the purpose is ―not to give [the text] a (more or less justified, more or less free)

meaning, but on the contrary to appreciate what plural constitutes it.‖137 Barthes continues to

develop the concept of the ideal text, writing that ―the networks are many and interact, without

any one of them being able to surpass the rest; this text is a galaxy of signifiers, not a structure of

signifieds; it has no beginning; it is reversible; we gain access to it by several entrances, none of

which can be authoritatively declared to be the main one ….‖ 138 In his study on the transmission

of the Tipiṭaka through the perspective of writing and orality in northern Thailand, Daniel

Veidlinger explains that ―when looking at the ‗roles‘ that manuscripts in particular have played,

it is essential to realize that manuscripts can fit into the lived practice of religious communities in

a variety of ways beyond their obvious function as support for the words of texts.‖139 And as

Payne has noted, it is impossible to characterize Buddhism as employing just one view of

language‘s potential. 140 It is the plurality and flexibility of texts which make them distinctively

suitable for artistic manipulation.

The visual manifestations of text not only reflect already established meanings, but also

create new interpretations of the signified and the nature and plurality of texts. Eagleton asserts,

136
Rambelli, Buddhist Materiality, 88-90.
137
Barthes, S/Z, 5.
138
Ibid.
139
Veidlinger, Spreading the Dhamma, 5.
140
Payne, ―Awakening and Language,‖ 89.

253
―every reading is always a rewriting;‖ 141 and every visual manifestation expounds and explores

the possibilities of sacred text, offering in its materiality new perspectives. The material

expression signifies as well as carries and communicates meaning, a concept as conveyed by

Henri-Jean Martin:

Writing systems are not disembodied, and written messages from past times are objects
that speak more than one language, dug out from the soil, discovered in tombs, or
transmitted from generation to generation, they often seem odd to us and a far cry from
our modern books by their very aspect they remind us that the shape of written signs
depends on the material on which they are written. When signs are written with care they
attest to an interest in proclamation and durability; when they are cursive they show that a
society was familiar with writing. When they are laid out without separations they remind
us that our modern page layouts are recent acquisitions. When they are written on scrolls
the text unfolds like a film. When only a small number of characters appears on each
page rapid reading proves impossible. Hence all these odd objects need careful scrutiny
before we can begin to understand what the always ambiguous relationship between
speech and text may have been in their own time. 142

In the jeweled-stūpa mandalas in particular, it is from the structure that the mandalas derive and

generate great significance and signification. The very materiality of texts is a signifier, so

ownership of material texts also carried great social and authoritative value. Veidlinger discusses

the idea of ‗metatextual‘ features of texts, proposing that outside of the text‘s traditional

components are features that likewise communicate a great deal, such as marginal writings,

corrections, calligraphic quality, fabric and many other features sometimes overlooked. 143 The

ubiquitous practice of shōgon 荘厳, or elaborate adornment of Buddhist visual culture and

architecture, stresses the importance of the materiality. Expensive and laborious commission can

signify a desire to not only manifest extreme piety but also wealth and social prestige. With the

jeweled-stūpa mandalas, beautifully dyed blue paper sets an exquisite background upon which

golden characters erect the central icon. Narrative images of gold and silver—and bright reds,

141
Eagleton, Literary Theory, 12.
142
Henri-Jean Martin, The History and Power of Writing, trans. Lydia G. Cochrane (Chicago and London:
University of Chicago Press, 1994), 43.
143
Veidlinger, Spreading the Dhamma, 103.

254
greens, blues, and yellows in the case of the Chūsonji set—surround the dharma reliquary. The

large scale of the project, eight hanging scrolls in the Ryūhonji set and ten in the Chūsonji and

Danzan Shrine versions, speaks to the costliness of the commissions. Indeed the rather large size

of each mandala only increases the significance of a project in which materiality is stressed.

Shōgon as well as the pattern of replication are popular means of generating tremendous amounts

of merit. Thus it stands to reason that the augmented embodiment of Buddha body, as both sūtra

and stūpa, and the manifestation of the Lotus Sūtra in word and image serve to amplify the

efficacious qualities of the scripture, providing multiple outlets to access the salvific potential of

the sūtra.

While Saussure ignored the material and historical aspects of signs, privileging instead

spoken word, later theorists have reclaimed the importance of materiality. 144 But rather than see

the material and oral expression of signs as two genres without overlap, Ruth Finnegan suggests

that written and oral manifestations are not rigid categories, but are often genres with permeable

borders.145 Numerous scholars have undertaken to flesh out the oral and aural qualities of sacred

word. Rambelli notes that when medieval texts were read, they were done so aloud and thus the

orality of texts is a critical component of ‗medieval textuality.‘ 146 William Graham, in his study

Beyond Written Word: Oral Aspects of Scripture in the History of Religion, advocates the

―fundamental orality of scripture,‖ or what he describes as the sensual dimension of text.147 Mary

Carruthers in her analysis on memory and texts in medieval cultures explains, ―A book is not

necessarily the same things as a text. ‗Texts‘ are the material out of which human beings make

‗literature.‘ For us, texts only come in books, and so the distinction between the two is blurred

144
For instance, Derrida challenges the privileging of orality over materiality. See Derrida, Of Grammatology.
145
Ruth Finnegan, Oral Poetry (Cambridge: Cambridge University Press, 1977), 16-24.
146
Rambelli, ―Texts, Talismans, and Jewels,‖ 55.
147
Graham, Beyond Written Word, ix.

255
and even lost. But, in a memorial culture, a ‗book‘ is only one way among several to remember a

‗text,‘ to provision and cue one‘s memory with ‗dicta et facta memorabilia.‘‖ 148 And while the

art historical approach of this project stresses the material expression of the sūtras composing,

both textually and pictorially, the jeweled-stūpa mandalas and the culture of veneration around

sacred manuscripts, it is important to consider that even though a fecundity of written texts

survive, this was not the primary method of textual participation and interaction in medieval

Japan.149 Because of this, it is interesting to consider the functions of writing and reading as well

as to remember the importance of the material expression of texts in such a context.

In her dissertation, ―Rendering the Body Buddhist: Sermonizing in Medieval Japan,‖

Charlotte Eubanks discusses the different forms of writing, from the very literal brush upon

paper to the metaphorical inscription of sacred word onto the mind and heart. She embraces a

broad definition and adopts Mary Carruthers‘ understanding that: ―Writing, then, is ‗anything

that encodes information in order to stimulate memory to store or retrieve information.‖ 150 In his

dissertation, ―Setsuwa, Knowledge, and the Culture of Reading and Writing in Medieval Japan,‖

Thomas Howell advocates a conventionally bifurcated understanding of reading: offertory and

interpretative. 151 Offertory reading is the oral performance of texts as ritual worship and does not

oblige hermeneutical understanding, 152 whereas interpretative reading is the exegetical

examination of texts for their substantive meaning. The second definition proposed by Howell

148
Mary Carruthers, The Book of Memory: A Study of Memory in Medieval Culture, 2nd ed. (Cambridge: Cambridge
University Press, 2008), 9-10.
149
In a later work on orality, Finnegan cautions against the tendency in scholarship to privilege written forms of text
over oral transmission. See Ruth Finnegan, Literary and Orality: Studies in the Technology of Communication
(Oxford: Basil Blackwell, 1988), 124-26.
150
Eubanks, ―Rendering the Body Buddhist,‖ 304. Carruthers, The Book of Memory, 8.
151
Howell, ―Setsuwa, Knowledge, and the Culture of Reading and Writing in Medieval Japan,‖ 172.
152
For more on the orality of text, see Shimizu Masumi‘s study on sūtra recitation, in which he tackles such issues
as the qualities of sound, the history of the practice, and the power generated from proper sūtra recitation. Shimizu
Masumi 清水眞澄, Dokyō no sekai nōdoku no tanjō 読経の世界—能読の誕生 (Tokyo: Yoshikawa Kōbunkan,
2001).

256
corresponds to Carruthers‘ emphasis on the memorial nature of reading: ―Reading is to be

digested, to be ruminated, like a cow chewing her cud, or like a bee making honey from the

nectar of flowers.‖153 She insists that a work ―is not truly read until one has made it part of

oneself—that process constitutes a necessary stage of its textualization merely running one‘s

eyes over the written pages is not reading at all, for the writing must be transferred into memory,

from graphemes on parchment or papyrus or paper to images written in one‘s brain by emotion

and sense.‖154 But as Rambelli writes, ―medieval religious texts were not necessarily and not

only ‗read,‘ and ‗reading‘ was not always and necessarily a personal, solitary and introspective

activity of disembodied decoding of inherent meaning of a text.‖155 Indeed some texts, called

hidensho 秘伝書 (‗hidden texts‘), were never even meant to be read but instead to be passed

down in secret boxes from one abbot to another.156 The esoteric reading of texts is a complex

process in which the content of the written word must be ―transposed through ritual to the

experiential realm of practice. Esoteric texts are to be grasped not through intellectual operations

alone but through and somatic exercises.‖157 As discussed in chapter four on the power of

dharma relics, Sasaki Kōkan explains that access to this power is achieved through numerous

endeavors, the reading of the sūtra being a most effective method. The particular technique of

tendoku 轉讀, whose general meaning is to chant the sūtra but usually refers to briefly chanting

the title, along with selected lines of text taken from portions of the scripture, certainly does not

involve a sustained nor deep engagement with the full text of the sūtra, but is nonetheless

153
Carruthers, The Book of Memory, 205.
154
Ibid., 11.
155
Rambelli, ―Texts, Talismans, and Jewels,‖ 52.
156
Rambelli, Buddhist Materiality, 92.
157
Abé, The Weaving of Mantra, 12.

257
incredibly potent.158 An even more abbreviated method of reading or more appropriately of

ritualistically handling the sūtra is that of the tenpon 轉翻, an active process involving holding

the text with both hands and moving it in a motion that mimics the flapping of a bird‘s wings

three times to the right, three times to the left, and once more in front. This dynamic treatment of

the sūtra occurs usually during chants of the sūtra.159 These abbreviated techniques are all in

great contrast to the ‗true reading‘ of the sūtra (真讀 shindoku) in which the full scripture is read.

As I described above, images are read as well. Optical registering of graphic images

suggests a visual mode of reading—one in which the viewer processes the graphic image for its

interconnected parts. This is particularly true in the case of images composed of or dominated by

text in which it is possible to not only read the graphic components of the painting, but also the

textual ones. Such visual reading of paintings infused with dharma relics allows for the

cognizance of sacred word‘s power. Interestingly, aside from reading as cognition and assembly

of meanings in paintings, motifs frequently found in texts such as illustrated sūtras and poetry

compilations can be read not only for their symbolic meaning but also for their phonetic value.

For example, a partially submerged, broken wheel can be read as the hiragana character, wa わ,

because of the similarities in form of the letter and the wheel. 160 Such occurrences can be called

phonograms (表音文字 hyōonmoji); however, because contextual understanding of the cultural

and religious references is required in order to read the pictures, the use of motifs in this context

might be better classified as ideograms (表意文字 hyōimoji).161 Komatsu Shigemi provides

further examples of these ideograms, such as the reading of ko こ as small baskets because of the

158
Sasaki, ―Sō no jushika to ō no saishika,‖ 53. For more on tendoku, see Shimizu Masumi, ―Nōdoku to nōsetsu:
ongei ‗dokyō‘ no ryōiki to tenkai 能読と能説: 音芸‗読経‘の領域と展開,‖ Ryōjin kenkyū to shiryō 梁塵: 研究と
資料 15 (1997): 25-29.
159
Sasaki, ―Sō no jushika to ō no saishika,‖ 52.
160
For good illustrations of this, see Komatsu, Heike nōkyō no kenkyū, 827.
161
Shibuya Kuritsu Shōto Bijutsukan, Mojie to emoji, 128.

258
small, circular-like shape of ko which mirrors the form of baskets. 162 Reading in the jeweled-

stūpa mandalas requires a negotiated viewing experience when interpreting the textual stūpa,

oscillating between imbricated image and text. The surrounding narrative vignettes require what

Gombrich has characterized as the decipherment of ―pictorial language.‖ 163 The illustrations, no

matter how conventional and rehearsed, necessitate a reading for their content, a critical merit-

generating practice known as kaisetsu 解説, which would provide the likely audience of

aristocrats well-versed in the Buddhist scriptures an opportunity to identify the passages from

which the vignettes derive. 164 And if the content of the illustration was considered ambiguous,

cartouches labeling most of the narratives would indicate the scriptural reference.

Conclusion

Within the larger project analyzing the jeweled-stūpa mandalas, this chapter explored the

complex relationships of text and image demanded by the mandalas. In order to demonstrate the

textualized community out of which the mandalas developed, I discussed a range of images

exhibiting innovative manipulations of text and image interactions. The conventional functions

of written word and graphic picture enacted in the mandalas were then revealed to be role

reversals. The jeweled-stūpa mandalas, as expressions of the non-hermeneutical, flexible uses of

scripture offer new perspectives on the potentialities of text and image. The sumptuous

materiality of the paintings, while certainly striking, also conveys the extreme piety, wealth, and

social prestige of the patron. The materiality of the text and consequently the sūtra‘s inventive

manipulation into a textual reliquary surrounded by narrative vignettes provides a variety of

162
Komatsu, Heike nōkyō no kenkyū, 828. Komastu provides many more examples, see Ibid., 823-29.
163
Ernst Gombrich, Art and Illusion: A Study in the Psychology of Pictorial Representation (London: Phaidon 1960),
9.
164
Sato Shinji, ed., Kenrantaru kyōten 絢爛たる経典 (Tokyo: Heibonsha, 1983), 97.

259
reading experiences for viewers. Overall, this chapter examined the visual manifestations of

scripture, for it is precisely the material expression of text‘s vast possibilities and the

composition of the mandalas that manifest the salvific matrix of sacred text and body.

260
 Chapter Seven

Conclusion

Goals and Summary of the Project

This dissertation set out to explore the practical and conceptual implications of the twelfth- and

thirteenth-centuries Japanese jeweled-stūpa mandalas. The mandalas are admittedly complex

paintings, but I believe that treating them as one of the most involved examples from the

eleventh through thirteenth centuries of innovative elaborations on sūtra transcription is the key

to unlocking their meaning. The project proceeded from a methodology grounded in visual

analysis and religious studies. I began with questions of semiotic inquiry about the prominence

and privileging of sacred text in the form of the central dharma reliquary, a characteristic

distinguishing the mandalas from nearly all other paintings before them. I sought to understand

the reasons behind the privileging of scripture on the picture plane and the inventive

manipulation of the sūtra text into the form of a stūpa, both novel choices in the context of their

early medieval Japanese production. In order to tackle these topics, the dissertation opened with

two chapters examining the practical issues concerning the mandalas such as questions of origins,

histories of the paintings, and a formal analysis of the compositions. Three subsequent chapters

explored the theoretical implications of the imbricated textual reliquary, all of which were based

on the fundamental issue of the conflation of text and image in the central dharma reliquary.

261
Method of Analysis

As the first extensive study of the paintings in English, I began with the practical issues raised by

the jeweled-stūpa mandalas. Motivated by the question of origins, I opened with an examination

of the continental prototypes of the mandalas. This exploration revealed the practice of textual

stūpa transcription as early as the tenth century in China. The earliest example of the textual

stūpa format exposed an interesting component not found in extant Japanese jeweled-stūpa

mandalas: the textual stūpa made from the Heart Sūtra manifested a strong dimension of puzzle-

solving. The characters of the sūtra do not proceed in an easily discernable path but rather

compel the viewer to recall the exact order of the scripture in order to solve the complicated

puzzle that twisters and maneuvers, overlapping itself repeatedly. The story of a twelfth-century

example displays an interest in theatricality; Fahui, a pious monk from late eleventh/twelfth-

century China, crafted an impressive three-dimensional textual stūpa whose characters soared

about the room when exposed to light for the audience of the emperor. The practice of textual

stūpa transcription in China continued through the twentieth century; however, much like the

extant Korean and Japanese examples, the records of the Qing textual stūpas do not demonstrate

a puzzle or theatrical component. Another crucial difference between the Chinese and Korean

examples and the Japanese jeweled-stūpa mandalas is that the Japanese paintings exhibit

narrative vignettes, adapted from the tales of the sūtra, encircling the central dharma reliquary.

Most likely prints of textual stūpas from either China or Korea made their way to Japan, where

they met with the culture of copying of the eleventh through thirteenth centuries and were

transformed into the Japanese jeweled-stūpa mandalas.

Having established the continental prototypes of the mandalas, I sought to uncover the

circumstances of sūtra transcription contemporary to the production of the jeweled-stūpa

262
mandalas. The conventional format for decorated scriptures took the form of blue and gold

illuminated scriptures of seventeen character lines. However, the culture of sūtra copying of the

eleventh through thirteenth centuries manifested a drive toward novelty demonstrated by the

artistic innovations in decorated sūtras. The origins of decorated sūtra copies were shown to

reach back to the eighth century, with various colors used for sūtra paper, gold and silver ink for

the script, and illustrated examples like the Illustrated Scripture of Cause and Effect. The

development of sūtra art in the eleventh through thirteenth centuries was one of sheer quantity

and innovative elaborations upon previously established decorative themes. Early medieval

scrolls, such as the Heike nōkyō and the Kunōjikyō, revealed sumptuous copying projects

involving multiple participants and resulting in large productions of opulent sūtra copies. The

structural divide between text and image, which assigns picture to the frontispiece of the scroll

and word to the subsequent lengths, began to break down at this time. Scrolls such as the Ichiji

butsu hokekyō (one character, Buddha Lotus Sūtra scroll), Ichiji hōtō hokekyō (one character,

jeweled-stūpa Lotus Sūtra scroll), and Ichiji rendai hokekyō (one character, lotus pedestal Lotus

Sūtra scroll) pair the sacred characters with accompaniments such as adjacently seated Buddhas,

enshrining stūpas, and supporting lotus pedestals, not only bridging the chasm between text and

image but also embodying the concept of dharma relics. The Lotus Sūtra fans and booklets as

well as the Menashikyō (literally, the ―eyeless sūtra‖) demonstrate a further deterioration of the

barrier by layering sacred text with mundane images of society.

Sūtra art was not the only area of innovation; sūtra transcription practices during the

eleventh through thirteenth centuries became more extreme. An examination of early medieval

documents uncovered the increasingly elaborate and intense copying practices of this period.

This trend was manifested in terms of quantity, such as group and individual projects

263
transcribing the entire Buddhist canon—sometimes in the lavish blue and gold sūtra format.

Particularly impressive examples come from Fujiwara Sadanobu (twelfth century) and the

mendicant monk, Shikijō (twelfth through thirteenth century), both of whom took decades of

resolute determination to complete the immense task of copying the Buddhist canon by hand.

Another indicator of the trend toward the extreme was the pace at which scriptures were

transcribed. For instance, the feat of copying the complete scriptural canon in one day

exemplified the extreme forms of transcription practice. Genuflection in the form of ichiji sanrei,

which required the copyist to pay observances (usually in the form of three bows after writing

each character), was characterized by attempts to incorporate more elaborate and extreme

devotional methods into copying the sūtras. The famous thirteenth-century Buddhist sculptor,

Unkei of the Kei school, and Madenokōji Nobufusa, a thirteenth/early fourteenth-century

courtier, both practiced this form of laborious transcription. The last form of extreme copying I

examined was the incorporation of alternative media, such as transcribing on stone, tiles, and

with blood. Pursuing this line of investigation revealed the jeweled-stūpa mandalas‘ context of

production, which exposed the distinctive trend toward innovation and extremism in eleventh-

through thirteenth-centuries sūtra art and associated religious practices. The mandalas were thus

discovered to be an iteration of this search for inventiveness characterizing the culture of copying

at this time.

The second investigation into the practical issues surrounding the jeweled-stūpa mandalas

explored the particular histories of each mandala set, along with the two independent mandalas

estranged from their original commissions, and conducted a formal analysis of the paintings. I

began with an examination of the mandalas from Chūsonji because more is known about this set

than the others. Unfortunately, the jeweled-stūpa mandalas of Chūsonji are not mentioned in any

264
early medieval texts, a curious state since much of the Ōshū Fujiwara‘s other commissions are

recorded. Therefore, an examination into the histories and commissions of the three generations

of rulers, combined with a comparative formal analysis of the extant Hiraizumi sūtra art and the

jeweled-stūpa mandalas, allowed for a general dating of the mandalas and the proffering of a

possible patron. As Miya Tsugio has pointed out, the extravagance of the mandalas warranted a

significant occasion to commemorate. Based on this style of analysis, most scholars have

concluded that Fujiwara Kiyohira was not the patron primarily because it is thought that if he had

commissioned the paintings then it would be documented in the pledge recording a massive

ceremony he held in 1126, and also because the stylistic characteristics of Kiyohira‘s blue and

gold illuminated Buddhist canon differ so drastically from the narrative vignette style of the

mandalas. Some scholars argue that Kiyohira‘s son, Fujiwara Motohira, was the patron; however,

too little remains, both in terms of records and sūtra art, to build a solid argument for Motohira. I

concluded that Hidehira was the most likely patron of the mandalas for a few reasons. The extant

sūtra art commissioned by Hidehira more closely matches the style of the narrative vignettes. In

1170, he was promoted to the constabulary position of ‗pacification‘ general. The appointment

ceremony was held at the imperial palace during the annual saishōkō, an imperially sanctioned

ceremony reaffirming the Golden Light Sūtra as guardian of the nation and legitimizer of

imperial authority. Thus, the conjunction between this important point in Hidehira‘s life and the

commission of the Golden Light Sūtra jeweled-stūpa mandalas revealed the paintings as a

symbolic monument to the power and authority of the Ōshū Fujiwara. The examination of the

formal qualities of the mandalas demonstrated the intimate and localized focus of the narrative

vignettes, exposing Hidehira‘s ambitions and anxieties as well as the emphasis on regional

religious faith.

265
 The same lacuna of information in early medieval records haunts the Danzan Shrine

jeweled-stūpa mandalas. However, by analyzing the formal qualities of the mandalas, I proposed

a late twelfth-century date based on the more sinicized style of the dense composition and

arrangement of the narrative vignettes as well as the brushwork, which is stiffer than that of the

eleventh century. This approximate date was further corroborated by an ink inscription on the

box housing the mandalas that ambiguously mentions a temple near-by called Shigaiji. It is

believed that this mortuary temple, founded in 1187 in honor of the Tendai monk Zōga,

commissioned the sumptuous set of jeweled-stūpa mandalas to commemorate Zōga‘s devotion to

the Lotus Sūtra and to dedicate to the monk the merit generated by the paintings‘ combination of

stūpa and sūtra. A formal analysis of the narrative vignettes revealed a more systematic

composition than discovered in the other sets. A few rules dictate the typical arrangement of the

narrative vignettes: paradisiacal scenes are grouped toward the upper half of the composition

while images of hell are assigned to the bottom of the scroll; chapter scenes are roughly grouped

together; and the scenes usually start in the lower left under the stūpa, moving upward to the top

of the narrative band before crossing over to the right side and continuing downward to the

concluding scene at the bottom right of the stūpa—a pattern seen in the Lotus Sūtra

transformation tableaux of Dunhuang.

The Ryūhonji mandalas present yet another elusive inscription hinting at their history. On

the backs of each painting, a black ink inscription records the mandalas‘ location in Hōryūji at

the time of the set‘s first recorded restoration in 1362. Internal records on the collection of

Hōryūji place the mandalas at the temple until the mid-sixteenth century when they were either

sold or gifted to Ryūhonji. The restoration of 1362 is of further significance to dating the

mandalas because Miya argues that when most paintings were restored, this occurred within the

266
first hundred years of their creation. Thus, an approximate mid-thirteenth-century date for the

Ryūhonji set is offered. Beyond these tantalizing clues, little else can be constructed of the

mandala‘s histories and movements. A formal analysis of the paintings also substantiated a date

later than the other two sets. The tidy composition and reduced number of scenes, along with the

firmness of the sūtra characters of the textual stūpa and the stricter brushwork of the narrative

vignettes, suggest a thirteenth-century date.

The two jeweled-stūpa mandalas separated from their original sets are now stored in a

private collection and at Myōhōji in Sakai. Beyond the clear connection with another lone

mandala housed at Jōshinji in Shiga prefecture strongly suggesting that the two originated from

the same commission, nothing more is known about the history of the privately owned mandala‘s

production or travels. As a result of restoration, the privately owned mandala features a brilliant

gold central stūpa, but the narrative vignettes are barely visible in most parts. The stūpa, which

radiates out from the darkened and largely indistinct background, lacks the decorative flourishes

seen in the other sets. The softness and roundness discernible in the treatment of the narrative

vignettes‘ figures reflects a style common to the late eleventh and early twelfth century, making

this painting the oldest example of the Japanese jeweled-stūpa mandalas. The lone mandala at

Myōhōji is in a far better state of preservation, but once again, the circumstances of the original

commission and how it came to be in the collection of Myōhōji eludes us. The mandala more

closely resembles the paintings of the Ryūhonji set; and given similarities such as the

architectural style of the stūpa, the selection and arrangement of the narratives, and the mirroring

of the cartouches, corroborated by the late twelfth-century style of the paintings reflected in the

brushwork of the landscape and the gentle facial expressions of the figures, it has been proposed

that the Myōhōji mandala was used as a model for the painters and copyists of the Ryūhonji set.

267
This places the production of the mandala sometime in the late twelfth century or early thirteenth

century. This second examination of the practical matters of the jeweled-stūpa mandalas

established the specific histories of the paintings, revealing their dating, contexts of their

particular commissions, and lives after production, using inscriptions and formal analyses. While

the first inquiry into the practical issues ascertained the prototypes of the mandalas and placed

the paintings within a contemporaneous and broader system of inventive transcription, this

investigation fleshed out the specific practical aspects of the jeweled-stūpa mandalas in

preparation for the theoretical exploration.

Building upon this practical and contextual foundation, the remainder of the dissertation

examined the jeweled-stūpa mandalas via readings into the meanings of the dharma reliquary.

These theoretical investigations focused on the implications of the imbricated central stūpa and

the relationships of text and image in that conflated icon. I first explored the reasons behind the

prominence and privileging of sūtra text, concluding that the power of sacred word compelled

the central role of scripture in the paintings. This power is usually indirectly proclaimed in a

variety of early medieval sources, so I examined claims in sūtras, setsuwa, and ecclesiastical

commentaries that sacred word is endowed with an active, salvific force. Sūtras routinely

champion sacred text as a great repository of dynamic power, and recording the innumerable

testaments far exceeds the boundaries of this project. Therefore, I chose to highlight the adamant

proclamations in the Lotus Sūtra, the Golden Light Sūtra, and the Perfection of Wisdom Sūtras as

an indicative sample of such widespread claims. The testaments within these scriptures

demonstrate the doctrinal justification for the power invested in sacred text. I then looked to

setsuwa and ecclesiastical commentaries to determine how this power proclaimed by the sūtras

was manifested and adopted. Several anthologies of setsuwa were analyzed, revealing the ability

268
of scripture to not only generate merit but also transform the world, saving lives and redeeming

souls. The open-ended nature of text and its limitless potential encourages the manifestation of

sūtra in various ways within Buddhist visual culture and religious practice. Because the objective

of this discussion was to demonstrate the early medieval attitude toward the power of scripture, I

analyzed a variety of commentaries from different schools of Buddhism which speak to the great

potential of sacred word. Given the influential nature of his writings as well as his strong

explanations of the potent power invested in scripture, I featured Kūkai‘s commentaries, which

advocate the possibilities of language in the process of enlightenment and reveal sacred word to

be open, all-inclusive texts. Early medieval Tendai texts and commentaries by Nichiren were

also examined for their proclamations on the abilities of scripture and language. After surveying

the wide-range of testaments to the power of sūtra, I concluded that the source of this efficacy

comes from the nonduality of the Buddha with his teachings, manifested as dharma relics. I

examined the long-attested notion of nonduality of the Buddha with the dharma, as established in

early Buddhist texts. This raised questions of what constitutes a dharma relic and when scriptural

text is considered a relic of the Buddha. I argued that it is through the proper veneration of the

sūtras as sacred, empowered objects through ritualistic preparation, worship, and visual

manifestations—such as in the elaborate scripture transcription in the jeweled-stūpa mandalas—

that the transubstantiation of paper into relic occurs and sūtra is revealed to be sacred and

powerful embodiments of the Buddha. By analyzing the religious practice of sūtra copying,

including the copying sūtras in accordance with ritual prescriptions, the burial of sacred text, and

the visual treatment of the sacred characters in sūtra copies, I contribute to the appraisal of sūtras

as dharma relics. The jeweled-stūpa mandalas were then shown to be embodiments of the great

reverence for and need to manifest the power of scripture, thus establishing a karmic connection

269
with this repository. Because the mandalas present an imbricated central icon, the merits

generated from copying sūtras and constructing stūpas are attained in one project. Therefore, the

mandalas are properly understood as visual manifestations of the conflation of the cult of relics

and the cult of the book, and as a reflection of the blended religious practices of medieval Japan.

I followed the investigation of the privileging of sūtra text with an examination of the

imbricated textual reliquary by exploring why the stūpa form was chosen, suggesting that the

imbricated icon embodied the multifaceted manifestations of the body of the Buddha in the form

of the dharma relics and the stūpa. This in turn called for an engagement with Buddha body

theory, in which I presented some ideas fundamental to the theory and introduced the three-body

system. Crucial to this study was the revelation that dharma relics are a manifestation of the

dharmakāya. I then examined the notions of the bodies of the Buddha as they are discussed in

the Lotus Sūtra and Golden Light Sūtra, since these scriptures were chosen to construct the

jeweled-stūpa mandalas. The Lotus Sūtra routinely identifies itself as nondual with the Buddha

and encourages its own enshrinement as a relic within stūpas. Importantly, the sūtra also claims

stūpas to be another manifestation of the Buddha‘s body. The Golden Light Sūtra similarly

claims that the Buddha‘s body is made of the dharma which is eternal.

Having established dharma relics as emanations of the dharmakāya of the Buddha, I

turned to the issue of the stūpa form as a manifestation of the Buddha‘s body, thereby explaining

the choice of the stūpa as a result not only of its meritorious and long-standing combination with

sūtra but, most importantly, because of its understanding as a body of the Buddha. Doctrinal

assertions claiming stūpas to be emanations of the body of the post- parinirvāṇa Buddha in the

form of the dharmakāya are corroborated by religious practices that treated the stūpa as the body

of the Buddha. Early Indian Buddhist art demonstrated this long-standing understanding of the

270
stūpa. Japanese early medieval visual culture also reflected this conflation of the stūpa with the

body of the Buddha. Beyond the somatic identifications of the stūpa, the salvific power of the

reliquary is yet another reason for its choice as the form sacred text assumes. This is

demonstrated by ad sanctos (‗at the place of saints‘) burials, begun in India and continued in

early medieval Japan, as seen in the burial practices on Mt. Kōya around the grave of Kūkai.

Numerous examples testify to the protective power of the stūpa. A legend explaining the

mystical origins of Shingon Buddhism champions the monument as the guardian of the teachings.

The salvific potentialities of the stūpa are also advocated in the tenth-century text, The Three

Jewels. I then showed that this faith in the stūpa was indicated by the many stūpa commissions

occurring in early medieval Japan, further compelled by the numerous injunctions in scripture to

construct reliquaries.

I concluded this examination by drawing together the implications of the discussions of

dharma relics and Buddha body theory for the jeweled stūpa mandalas. By showing the

imbricated textual stūpa to be the locus of the somatic and textual web revealing the conflation of

sūtra, dharma, body, relic, and stūpa in doctrine and praxis, I explicated the main unifying

principle underpinning the jeweled-stūpa mandala‘s construction as the visual locus of a salvific

matrix of text and body. It is through the inventive manipulation of text that the bodies of the

Buddha are fashioned as an image composed of sūtra and presenting the conceptual

representations of the Buddha‘s body. In the jeweled-stūpa mandalas, the complete imbrication

of both materializations of the dharmakāya serves as a visual treatise on the ultimate

indivisibility of body: it is unfeasible to behold the stūpa without reading it as the sūtra, and it is

impossible to see the sūtra without regarding the stūpa. The conceptual threads of text, dharma,

body, relic, and stūpa at work in the mandalas are utterly interwoven; pulling any one thread

271
inevitably leads to an intersection joining other strands of the salvific matrix. Indeed, I argue that

the inextricably imbricated nature of these threads and the visual manifestation of that conflation

in the textual reliquary of the jeweled-stūpa mandalas serve as the very basis of the paintings.

The final theoretical examination explored the nature of text and image relationships in

the mandalas from an analytical, visual perspective based in semiotics. Although for practical

reasons this was the first explicit discussion of semiotics, the entire dissertation is based in

semiotic concerns over the nature of representation, construction of meaning, and the function of

text. After arguing for the place of semiotic-informed analysis in the examination of early

medieval Buddhist visual culture in Japan and specifically in images exhibiting strong

interactions of word and picture, I examined the textualized community out of which the

innovative jeweled-stūpa mandalas emerged in order to establish the precedence of intriguing

text and image relationships. I began with images that layer the divine and profane such as

occurs in the Lotus Sūtra fans and booklets and the Menashikyō as well as shōsokukyō (personal

letters upon which are transcribed sacred scriptures) which are visually rich memorial

phenomena that have received little attention in art history. While these examples certainly span

the gap conventionally separating text and image and layer text upon image, ensuring that the

two are read together, word and picture nonetheless retain their own distinct visual properties.

The imbricated images of scrolls like the Heike nōkyō and the Collection of Japanese and

Chinese Verses [with ashide] display a puzzle-like play with word and picture reminiscent of the

early continental jeweled-stūpa mandala prototypes. In these images, textual ashide (―reed-hand

script‖) masquerade as graphic pictures. The kanji, kana, and hiragana characters elongate, twist,

and abbreviate into commonly found motifs such as rocks, birds, and reeds. In these ways, text

becomes mini-pictures within a larger graphic composition. I then moved further along the

272
spectrum of text and image collaborations with a discussion of what I termed ―empowered

inscriptions.‖ Nichiren‘s Great Mandala is a prime example of the replacement of image by text.

Graphic picture is utterly abandoned, and textual inscriptions ascend to the primary visual feature

of the mandalas. These mandalas represent yet another inventive twist in the relationships of text

and image, for word itself becomes picture. Establishing the spectrum of text and image

interactions in the eleventh- through thirteenth-century visual community exposed the singularity

of the relationships witnessed in the jeweled-stūpa mandalas.

I then discussed the role reversals of word and picture experienced in the mandalas.

Because of the manipulation of sūtra text into the form of a stūpa, the narrative vignettes must

shoulder the burden of imparting the scriptural stories. I analyzed the functions of the text that

have jettisoned their hermeneutical dimensions and conducted a reading of the narrative

vignettes through the use of specific scenes capturing the tales of the Medicine King Bodhisattva.

In the jeweled-stūpa mandalas, the textual stūpa becomes a graphic image both in terms of

function and appearance, and the narrative vignettes are read as content from the text. I explored

this particular type of text and image imbrication as it occurs in other countries over a great

sweep of time, offering a brief survey of this conflated relationship seen in Hebrew micrography,

Sufi art of Amadou Bamba, and the work of Xu Bing. The collusions of word and picture found

in the jeweled-stūpa mandalas and works from other cultures and religions speak to the power of

calligrams as explained by Foucault. The complexities of the text and image relationships in the

mandalas compel a discussion about ways of viewing. I adopted an explanation known as

―syncopated viewing‖ in which the imbricated interactions of word and picture compel the

viewer to constantly interchange vantage points in order to process the complex paintings. I also

explored the potential roles and limits of the basic semiotic components of signifier and signified.

273
The materiality of the jeweled-stūpa mandalas was then considered and revealed to be laden with

symbolism and the power of expression. But it was also important to consider the oral and aural

dimensions of early medieval texts as well as the different understandings of the process of

reading. By breaking down the interdependent components of the jeweled-stūpa mandalas‘

complex structure and showing how the paintings oblige engaged readings and viewings, this

final theoretical examination explicated how the complicated relationships between text and

image serve to construct meaning in the jeweled-stūpa mandalas.

Implications of the Dissertation

This study of the jeweled-stūpa mandalas revealed that although the paintings emerged for a

relatively short period of time during the twelfth and thirteenth centuries, they nonetheless

represent the pinnacle of a historical development characterized by the combination of

innovative trends in sūtra art and transcription practices and faith during the emergent time of

what was considered the latter days of the Law (mappō). By locating these mandalas within a

coherent milieu, the paintings themselves were made more intelligible and revealed to be a rare

yet important part of an interconnected system of art, praxis, and belief surrounding sacred text

in early medieval Japan. The study also gave fuller form to the understanding of the milieu itself

by examining and synthesizing the trends toward inventiveness in religious expression and

practice. Making the mandalas comprehensible in turn helped to elucidate the practices that went

into making them, practices which previously have been regarded as aberrations. Quite the

contrary, this dissertation showed that those religious practices and the consequent sūtra art were

no bastardizations of faith and its expression, but were exemplary of distinctive artistic pursuits

in the service of religious devotion.

274
 The novel structure of the jeweled-stūpa mandalas calls for an examination into the role

of dharma relics not only in the paintings, but within the context of the mandalas‘ early medieval

production. Sūtra text was revealed to play a crucial function in visual culture, praxis, and faith

as the object of inventive reverence during the eleventh through thirteenth centuries. The

imbrication of sūtra and stūpa in the dharma reliquary likewise demonstrates the importance of

Buddha body theory to art historical inquiry, and thus encourages a fuller consideration of this

challenging and engaging doctrinal aspect with respect to art. I used the concept of the salvific

matrix of text and body specifically to analyze the mandalas, but the concept is itself a reflection

of a broader phenomenon in early medieval religious doctrine and practice which often found

expression in visual culture. In the jeweled-stūpa mandalas, the complexity of the artistic web of

meaning reflects the complicated interrelations of the concepts themselves: isolating a single

thread is impossible, and without exploring the interconnections, understanding the mandalas

and the concepts so innovatively unified in the textual stūpa is impossible. The imbrication of

text, dharma, body, relic, and stūpa in the central icon of the mandalas signifies the same

conflation so often encountered in doctrine, praxis, and other examples of visual culture. In the

jeweled-stūpa mandalas, the form is the content. Therefore, far from being an anomaly flowing

from aberrant sūtra transcription practices of the eleventh through thirteenth centuries, the

mandalas represent an artistic apotheosis of the interconnected and ultimately inextricable

concepts of text, dharma, body, relic, and stūpa as expressed in doctrine and praxis. The

exploration of the fundamental components constructing the mandalas was based in the

semiotically informed question of how parts work together and what they signify individually

and collectively. Beginning with this point of entry retains a focus on how meaning is crafted

275
and, because of the complex structure, how viewers optically and cognitively confront the

jeweled-stūpa mandalas—foci without which the paintings cannot be adequately understood.

Avenues for Future Research

I hope to have demonstrated that the analyses applied to the study of the jeweled-stūpa mandalas

have wider implications beyond these paintings alone. The root of the study was an examination

of the meanings constructed by the text and image relationships, and this methodology can be

applied to other Japanese art historical investigations. In this project I attempted to discuss the

general kinds of word and picture interactions occurring in the early medieval period, but the

discussion was by no means comprehensive. Beyond more in-depth analyses of the text and

image implications of the secondary works already mentioned, many other examples are eligible

for consideration.

For instance, the popular category of seed-syllable mandalas (種字曼荼羅 shuji

mandara), in which the anthropomorphic form of the deities is replaced by their seed-syllable

characters (a script in Japan which uses Siddhaṃ to render Sanskrit syllables), are amenable to

the method of analysis pursued here.1 Seed-syllables were thought to embody the distillation of

the deity‘s essence. A further discussion of the power of word and the conventional reversal of

text‘s role in the paintings, along with a consideration of Buddha body could further illuminate

the meanings of the paintings.

Examples of other empowered inscriptions in which text is privileged, the category of

paintings known as myōgō honzon 名号本尊 (the name of a Buddha or a powerful verse that is

treated as an icon), would be another fruitful area of exploration. This format of painting

1
For an image, see Shibuya Kuritsu Shōto Bijutsukan, Mojie to emoji no keifu, 27 figs. I-5 and I-6.

276
continued from the twelfth century throughout the medieval period. The foundational myōgō

honzon is the nenbutsu (calling on the name of the Buddha), Namu amida butsu 南無阿弥陀仏

(homage to Amida Buddha), otherwise known as the rokuji myōgō 六字名号 or the six-character

formulation.2 Within this categorization of paintings, commonly associated with the Pure Land

schools, are several additional subvariants, including the slightly longer chants using different

names for Amida such as the kuji myōgō 九字名号 or nine-character formulation, Namu

fukashigikō nyorai 南無不可思議光如来 (homage to the unfathomable radiant Buddha), and the

jūji myōgo 十字名号 or ten-character formulation, Kimyō jin jippō mugekō nyorai 帰命尽十方

無碍光如来 (homage to the Buddha of unhindered light that illuminates the ten directions).3 Of

similar visual construction to the myōgō honzon are the kōmyō honzon 光明本尊 (sacred light

inscriptions). These images combine nenbutsu inscription with the patriarchal portrait tradition. 4

The typical structure of the kōmyō honzon places the nenbutsu inscription in the center of the

composition with rays of divine light emanating outward, embracing the surrounding portraits of

the patriarchs, such as the thirteenth-century example at Myōgenji 妙源寺 in Aiwa.5 The kōmyō

honzon offer another opportunity to examine the roles sacred inscription and graphic portraiture

play in artistic compositions.

Another type of early medieval imagery with interwoven text and image issues is the

imbricated painting known as the Sanzen butsymyō kyō hōtō zu 三千仏名経宝塔図 in the

collection of Fukaji 冨賀寺 in Aichi prefecture, which uses the text of the Sūtra of the Names of

2
For a Momoyama period example, see Ibid., 28 fig. I-9.
3
For a Muromachi period image of the ten-character formulation, see Ibid., 28 fig. I-8.
4
Andō Fumihito 安藤 章仁, ―Shinran no shinbutsu shosha seikyō ni tsuite 新発見の真佛書写聖教について,‖
Indogaku bukkyōgaku kenkyū 印度學佛教學研究 58, no. 2 (2010): 1097-92.
5
For several examples of kōmyō honzon, including the Myōgenji example, see Osaka Shiritsu Bijutsukan 大阪市立
美術館, ed., Shōtoku Taishi shinkō no bijutsu 聖徳太子信仰の美術 (Osaka: Tōhō Shuppan, 1996), 163-67 figs.
289-96.

277
Three Thousand Buddhas.6 This mid-fourteenth-century painting offers a very similar text and

image relationship as that seen in the jeweled-stūpa mandalas, particularly so because the

painting features a prominent stūpa also made of sūtra text. In this case, the names of the

Buddhas as recorded in the scripture construct the stūpa and the halo of emanating light. Not

only should the roles of word and picture be considered with this painting, the use of Buddha

body theory would likely be informative.

Of course, further studies on later manifestations of textual stūpas would benefit from the

analyses employed in this dissertation. Printed examples from the fifteenth and sixteenth

centuries of textual stūpas composed of Siddhaṃ-style letters would benefit from an analysis of

the somatic implications of the structure.7 A late eighteenth/nineteenth-century example of a

textual stūpa by the literatus, Tanomura Chikuden 田能村竹田 (1777-1835), while lacking the

narrative vignettes of other textual stūpas after the thirteenth century, resembles the construction

and arrangement of the jeweled-stūpa mandalas.8 Katō Nobukiyo 加藤信清 (1734-1810), a

samurai official and devout Buddhist, composed numerous paintings using sacred text to

construct the images. For example, Descent of Amida with Two Attendant Bodhisattvas painted

in 1796 uses the characters from three Pure Land scriptures to craft the entire composition. 9

From 1788-92, he also composed fifty paintings from a series on the five hundred arhats all

rendered with the text of the Lotus Sūtra.10 Katō‘s paintings offer a fascinating opportunity to

explore the issue of Buddha body and dharma relic conflation, principally so because sacred text

6
For an image, see Hayashi On, Kamakura bukkyō kaigakō 鎌倉仏教絵画考 (Tokyo: Chūō Kōron Bijutsu Shuppan,
2010), 67-69 figs. 48-51.
7
For examples, see Ishida, ―Tō, tōba, sutsūpa,‖ 90 figs. 244 and 245.
8
For images of this textual stūpa, see Ibid., 95 figs. 253 and 254.
9
For an image, see ―Katō Nobukiyo, Descent of the Buddha Amida with two attendant bodhisattvas, a hanging
scroll painting,‖ British Museum, accessed September 5, 2011,
http://www.britishmuseum.org/explore/highlights/highlight_objects/asia/k/kat%C5%8D_nobukiyo,_descent_of_the.
aspx.
10
For an image, see Shibuya Kuritsu Shōto Bijutsukan, Mojie to emoji no keifu, 55 fig. II-9.

278
creates the anthropomorphic body of deities. Other Edo period examples featuring an imbricated

text and image relationship are too copious to recount.11 Many of these Edo paintings and prints

evince a complete reversal of the conventional roles of text and image, and a semiotic

exploration of this phenomenon would shed light on the explosion of examples featuring

complicated word and picture relationships at this time.

Certainly, the applicability of Buddha body, dharma relic, and stūpa explorations, text

and image examinations, and the analysis of the power of sacred text is far broader than can be

discussed in this conclusion, and the examples given are meant only to present a sample of

different ways to adapt and expand the analyses used in this dissertation. Beyond the explication

of the jeweled-stūpa mandalas, the larger goal of this dissertation was to propose new ways of

expanding the approach of art historical analysis and to urge a sustained consideration of the

powerful relationships between text and image, because the implications of these relationships

can offer fuller understandings of not only visual culture but also the socio-religious milieu.

11
For plentiful examples, see Ibid.

279
 Bibliography

Abé, Ryūichi. The Weaving of Mantra Kūkai and the Construction of Esoteric Buddhist
Discourse. New York: Columbia University Press, 1999.
Akiyama Terukazu 秋山光和. ―Women Painters at the Heian Court.‖ Translated by Maribeth
Graybill. In Flowering in the Shadows: Women in the History of Chinese and Japanese
Painting, edited by Marsha Weidner, 159-184. Honolulu: University of Hawai‘i Press,
1990.
Akiyama Terukazu, Yanagisawa Taka 柳澤孝, and Suzuki Keizō 鈴木敬三. Senmen hokekyō no
kenkyū 扇面法華経の硏究. Tokyo: Kajima Shuppankai, 1972.
Andō Fumihito 安藤 章仁. ―Shinran no shinbutsu shosha seikyō ni tsuite 新発見の真佛書写聖
教について.‖ Indogaku bukkyōgaku kenkyū 印度學佛教學研究 58 no. 2 (2010): 1097-
92 [reverse pagination].
Ariga Yoshitaka 有賀祥隆. ―Hokekyō-e 法華経絵.‖ Nihon no bijutsu 269 (1988).
______. ―Konkōmyō saishōō kyō zu saikō 金光明最勝王経金字宝塔曼荼羅図再考.‖ Chūsonji
bukkyō bunka kenkyūjo ronshū 中尊寺仏教文化研究所論集 1 (1997): 92-99.
Austin, J. L. How to Do Things with Words. Oxford: Oxford University Press, 1976.
Avrin, Leila. Micrography as Art. Jerusalem: The Israel Museum, Jerusalem, 1981.
______. ―Hebrew Micrography: One Thousand Years of Art in Script.‖ Visible Language 18,
no.1 (1984): 87-95.
Bailey, D.R. Shackleton, ed. The Śatapañcāśatka of Mātṛceṭa. Cambridge: Syndics of the
Cambridge University Press, 1951.
Bakhtin, Mikhail. The Dialogic Imagination: Four Essays. Translated by Caryl Emerson and
Michael Holquist. Austin and London: University of Texas Press, 1981.
Bal, Mieke. Reading “Rembrandt” Beyond the Word-Image Opposition. Cambridge:
Cambridge University Press, 1991.
Barthes, Roland. S/Z. Translated by Richard Miller. Oxford: Blackwell, 1990.
Baxter, James C., and Joshua A. Fogel, eds. Writing Histories in Japan: Texts and Their
Transformations from Ancient Times through the Meiji Era. Kyoto: International
Research Center for Japanese Studies, 2007.
Beal, Samuel, trans. Si-yu-ki—Buddhist Records of the Western World: Translated from the
Chinese of Hiuen Tsiang (A. D. 629), vol. 2. London: Trübner & Co., 1884.
Belsey, Catherine. Critical Practice. London: Routledge, 2002 [1980].
Bendall, Cecil, and W.H.D. Rouse, trans. Śikshā-samuccaya: A Compendium of Buddhist
Doctrine. London: John Murray, 1971[1902].
Benn, James A. ―Where Text Meets Flesh: Burning the Body as an Apocryphal Practice in
Chinese Buddhism.‖ History of Religions 37 (1998): 295-322.
______. Burning for the Buddha: Self-Immolation in Chinese Buddhism. Honolulu: University of
Hawai‘i Press, 2007.
Boucher, Daniel. ―The Pratityasamutpadagatha and Its Role in the Medieval Cult of the Relics.‖
The Journal of the International Association of Buddhist Studies 14 (1993): 1-27.
______. ―Sūtra on the Merit of Bathing the Buddha.‖ In Buddhism in Practice, edited by Donald
S. Lopez, 59-68. Princeton: Princeton University Press, 1995.
British Museum. ―Katō Nobukiyo, Descent of the Buddha Amida with two attendant
bodhisattvas, a hanging scroll painting.‖ Accessed September 5, 2011.
280
 http://www.britishmuseum.org/explore/highlights/highlight_objects/asia/k/kat%C5%8D_
nobukiyo,_descent_of_the.aspx.
Bussho Kankōkai 佛書刊行會, ed. ―Nittō guhō junrei gōki 入唐求法巡礼行記.‖ In Dainihon
bukkyō zensho 大日本佛教全書, vol. 113, 169-282. Tokyo: Bussho kankōkai, 1912-1922.
Carruthers, Mary. The Book of Memory: A Study of Memory in Medieval Culture, 2nd ed.
Cambridge: Cambridge University Press, 2008.
Conze, Edward, trans. The Perfection of Wisdom in Eight Thousand Lines and Its Verse
Summary, 5th ed. San Francisco: Four Seasons Foundation, 1995 [1973].
______. Buddhist Wisdom: The Diamond Sutra and the Heart Sutra. Toronto: Random House,
2001[1958].
Davids, Mrs. Rhys. The Milinda-Questions: An Inquiry into Its Place in the History of Buddhism
with a Theory as to Its Author. London: Routledge, 2000[1930].
Davids, T.W. and C.A.F. Rhys, trans. Dialogues of the Buddha: Translated from the Pali of the
Dīgha Nikāya, Part II. London: Henry Frowde, Oxford University Press, 1938.
Davids, T.W. Rhys, and William Stede, eds. The Pali Text Society’s Pali-English Dictionary
Chipstead, Surrey: The Pali Text Society, 1925.
Demiéville, Paul. ―Busshin.‖ In Hōbōgirin, edited by Paul Demiéville, 174-85. Tokyo: Maison-
Franco-Japonaise, 1929.
Derrida, Jacques. Of Grammatology. Translated by Gayatri Chakravorty Spivak. Baltimore and
London: The Johns Hopkins University Press, 1976.
Digital Dictionary of Buddhism. Accessed May 10, 2010. http://buddhism-
dict.net.www2.lib.ku.edu:2048/cgi-bin/xpr-ddb.pl?q=銖.
Dobbins, James C. ―Portraits of Shinran in Medieval Pure Land Buddhism.‖ In Living Images:
Japanese Buddhist Icons in Context, edited by Robert H. Sharf and Elizabeth Horton
Sharf, 19-48. Stanford: Stanford University Press, 2001.
Donner, Neal, and Daniel B. Stevenson. The Great Calming and Contemplation: A Study and
Annotated Translation of Chih-i’s Mo-ho Chih-Kuan. Honolulu: University of Hawai‘i
Press, 1993.
Dykstra, Yoshiko Kurata, trans. Miraculous Tales of the Lotus Sutra from Ancient Japan: The
Dainihonkoku hokekyōkenki of Priest Chingen. Ōsaka: Intercultural Research Institute,
Kansai University of Foreign Studies, 1983.
Eagleton, Terry. Literary Theory: An Introduction. Oxford: Basil Blackwell, 1983.
Eckel, Malcolm David. To See the Buddha: A Philosopher's Quest for the Meaning of Emptiness.
Princeton: Princeton University Press, 1994.
Egami Yasushi 江上綏. ―Sōshokukyō 装飾経.‖ Nihon no bijutsu 日本の美術 278 (1989).
Emmerick, R.E., trans. The Sūtra of Golden Light, 3rd ed. Oxford: Pali Text Society, 1996.
Eubanks, Charlotte. ―Rendering the Body Buddhist: Sermonizing in Medieval Japan.‖ PhD diss.,
University of Colorado, Boulder, 2005.
Fabricand-Person, Nicole. ―Demonic Female Guardians of the Faith: The Fugen Jūrasetsunyo
Iconography in Japanese Buddhist Art.‖ In Engendering Faith: Women and Buddhism in
Premodern Japan, edited by Barbara Ruch, 343-82. Ann Arbor: Center for Japanese
Studies, University of Michigan, 2002.
Faure, Bernard. The Rhetoric of Immediacy: A Critical Critique of Chan/Zen Buddhism.
Princeton: Princeton University Press, 1991.
Ferber, Stanley. ―Micrography: A Jewish Art Form.‖ Journal of Jewish Art 3/4 (1977): 12-24.
Finnegan, Ruth. Oral Poetry. Cambridge: Cambridge University Press, 1977.

281
______. Literary and Orality: Studies in the Technology of Communication. Oxford: Basil
Blackwell, 1988.
Fontein, Jan. ―Śarīra Reliquary from Pagoda of Powŏn-sa Temple Site.‖ Misul charyo 47
(1991): 104-08.
Foucault, Michel. The Archaeology of Knowledge. Translated by A. M. Sheridan Smith. London:
Tavistock Publications, 1972.
______. The Order of Things: An Archaeology of the Human Sciences. New York: Vintage,
1973.
______. This Is Not a Pipe. Translated and edited by James Harkness. Berkeley; Los Angeles;
London: University of California Press, 1983.
Fowler, Sherry D. Murōji Rearranging Art and History at a Japanese Buddhist Temple.
Honolulu: University of Hawai‘i Press, 2005.
Fujita Tsuneyo 藤田経世, ed. ―Tōbōki 東宝記‖ In Kōkan bijutsu shiryō 校刊美術史料, vol. 2.
Tokyo: Chūō Kōron Bijutsu Shuppan, 1975.
______, ed. ―Sanbutsu jōshō 讃仏乗抄.‖ In Kōkan bijutsu shiryō 校刊美術史料, vol. 3. Tokyo:
Chūō Kōron Bijutsu Shuppan, 1975.
Fujiwara Munetada 藤原宗忠. ―Chūyūki 中右記.‖ In Zōhō shiryō taisei 増補史料大成, vol. 13.
Tokyo: Rinsen Shoten, 1965.
Fujiwara Sanesuke 藤原実資. Shōyūki 小右記. In Dai nihon kokiroku 大日本古記錄, vol. 6.
Tokyo: Iwanami Shoten, 1971.
Fujiwara Tametaka 藤原為隆. ―Eishōki 永昌記.‖ In Yōmei sōsho 陽明叢書, vol. 29, edited by
Yōmei Bunko 陽明文庫. Kyoto: Shibunkaku Shuppan, 1988.
Gandelman, Claude. ―By Way of Introduction: Inscriptions as Subversion.‖ Visible Language 23
no. 2/3 (1989): 141-169.
Gardiner, David. ―Kūkai's View of Exoteric Buddhism in his Benkenmitsu nikyōron.‖ Bulletin of
the Research Institute of Esoteric Buddhist Culture 5 (1992): 161-202.
Giles, Lionel. Descriptive Catalogue of the Chinese Manuscripts from Tunhuang in the British
Museum. London: Trustees of the British Museum, 1957.
Gombrich, Ernst. Art and Illusion: A Study in the Psychology of Pictorial Representation.
London: Phaidon 1960.
______. The Image and The Eye: Further Studies in the Psychology of Pictorial Representation.
Oxford: Phaidon, 1994.
Gómez, Luis O. ―The Whole Universe as a Sūtra.‖ In Buddhism in Practice, edited by Donald S.
Lopez, 107-112. Princeton: Princeton University Press, 1995.
Graham,William. Beyond Written Word: Oral Aspects of Scripture in the History of Religion.
Cambridge: Cambridge University Press, 1993.
Groner, Paul. Saichō The Establishment of the Japanese Tendai School. Honolulu: University of
Hawai‘i Press, 2000.
Grosnick, William H. ―The Tathāgatagarbha Sūtra.‖ In Buddhism in Practice, edited by Donald
S. Lopez, 92-106. Princeton: Princeton University Press, 1995.
Gu Yi 顧逸, ed. Xuanhe shupu 宣和書譜. Shanghai: Shanghai Shuhua Chubanshe, 1984.
Guang Xing. The Concept of the Buddha: Its Evolution from Early Buddhism to the Trikaya
Theory. London and New York: Routledge Curzon, 2005.
Guoli gugong bowuyuan 國立故宮博物院, ed. Midian Zhulin 秘殿珠林. Taibei: Guoli gugong
bowuyuan, 1971.

282
Habito, Ruben L. F. ―The Notion of Dharmakāya: A Study in the Buddhist Absolute.‖ Journal of
Dharma 11 (1986): 348-378.
______. ―Buddha-body Theory and the Lotus Sutra: Implications for Praxis.‖ In A Buddhist
Kaleidoscope: Essays on the Lotus Sutra, edited by Gene Reeves, 305-18. Tokyo: Kōsei
Publishing Company, 2002.
Haga Noboru 芳賀登. Sōgi no rekishi 葬儀の歴史. Tokyo: Yūzankaku 雄山閣, 1996.
Hakeda, Yoshito S., trans. Kūkai: Major Works. New York, Columbia University Press, 1972.
Hamada Takashi 浜田隆. ―Konkōmyōsaishōōkyō kinji hōtō mandara zu 金光明最勝王経金字
宝塔曼荼羅図.‖ In Chūsonji 中尊寺, edited by Fujishima Gaijirō 藤島亥治郎, 145-52,
261-65. Tokyo: Kawade Shobō Shinsha, 1971.
Haraguchi Shizuko 原口志津子. ―Toyama, Honpōjizō ‗Hokekyō mandara‘ no zuzō kaishaku to
kanjinsō Jōshin 富山, 本法寺蔵『法華経曼茶羅』の図像解釈と勧進僧浄信.‖ Kyōto
bigaku bijutsushigaku 京都美学美術史学 3 (2004): 27-66.
Harrison, Paul. ―Is the Dharma-kāya the Real ‗Phantom Body‘ of the Buddha?‖ Journal of the
International Association of Buddhist Studies 15, no. 1 (1992): 44-94.
Harvey, Peter. "The Symbolism of the Early Stūpa." Journal of the International Association of
Buddhist Studies 7, no. 2 (1984): 67-93.
Hayashi On 林温. ―Daichōjuinzō konkōmyōsaishōōkyō kinji hōtō mandara zu oboegaki 大長寿
院蔵金光明最勝王経金字宝塔曼荼羅図覚え書き.‖ Bukkyō geijutsu 仏教芸術 277
(2004): 81-95.
______. Kamakura bukkyō kaigakō 鎌倉仏教絵画考. Tokyo: Chūō Kōron Bijutsu Shuppan,
2010.
Hickman, Brian. ―A Note on the Hyakumantō Dhāranī.‖ Monumenta Nipponica 30 (1975): 87-
93.
Hiraizumi Chōshi Hensan Iinkai 平泉町史編纂委員会, ed. ―Chūsonji rakkei kuyō ganmon.‖ In
Hiraizumi chōshi 平泉町史, vol. 1, 59-61. Tokyo: Zoku Gunsho Ruiju Kanseikai, 1985.
Hopkins, Jeffrey, and Lati Rimpoche, with Anne Klein, eds. and trans. The Precious Garland
and The Song of the Four Mindfulnesses: Nāgārjuna and the Seventh Dalai Lama.
London: George Allen and Unwin Ltd., 1975.
Howell, Thomas R. ―Setsuwa, Knowledge, and the Cultures of Reading and Writing in Medieval
Japan.‖ PhD diss., University of Pennsylvania, 2002.
Huntington, Susan L. ―Early Buddhist Art and the Theory of Aniconism.‖ Art Journal 49 (1990):
401-408.
______. ―Aniconism and the Multivalence of Emblems: Another Look.‖ Ars Orientalis 22
(1992): 111-56.
Hurvitz, Leon, trans. Scripture of the Lotus Blossom of the Fine Dharma. New York: Columbia
University Press, 2009.
Hye-Bong Ch‘ŏn. ―Dhāraṇī-Sutra of the Early Koryŏ.‖ Korea Journal 12 (1972): 4-12.
Ikemi Sumitaka 池見澄隆. ―Tsumi to sono kaiketsu 罪とその解決.‖ In Hokekyō no shinri
sukui o motomete 法華経の真理: 救いをもとめて, edited by Miya Tsugio, 116-129.
Tokyo: Shūeisha, 1989.
Ikumi Kaminishi. Explaining Pictures: Buddhist Propaganda and Etoki Storytelling in Japan.
Honolulu: University of Hawai'i Press, 2006.
Ishida Mosaku 石田茂作. ―Kokuhō saishōōkyō kyōtō mandara 国宝最勝王経経塔曼荼羅.‖ In
Chūsonji okagami 中尊寺大鏡, vol. 2, 4-13. Tokyo: Ōtsuka Kōgeisha, 1941.
283
______, ed. Chūsonji 中尊寺. Tokyo: Asahi Shinbunsha, 1959.
______. ―Gangōji gokurakubō hakken no kokerakyō 元興寺極楽坊発見の杮経.‖ In Gangōji
gokurakubō chūsei shomin shinkō shiryō no kenkyū 元興寺極楽坊: 中世庶民信仰資料
の研究, edited by Gorai Shigeru 五来重, 229-38. Kyoto: Hōzōkan, 1964.
______. Nihon buttō, zuhan 日本佛塔, vol. 2. Tokyo: Kōdansha, 1969.
______. ―Tō, tōba, sutsūpa 塔 塔婆・スツーパ.‖ Nihon no bijutsu 日本の美術 77 (1972).
Iwate Nippōsha 岩手日報社, ed. Yomigaeru hihō Chūsonji konjikidō よみがえる秘宝: 中尊寺
金色堂. Morioka 盛岡: Iwate Nippōsha, 1974.
Izumi Takeo 泉武夫. ―Hokekyō hōtō mandara 法華経宝塔曼荼羅.‖ Kokka 国華 1169 (1993):
29-38.
Japanese Architecture and Art Net Users System.‖ Kochōsō.‖ Accessed December 1, 2010.
http://www.aisf.or.jp/~jaanus/.
Kageyama Haruki 景山春樹. ―Yokogawa ni okeru nyohō shakyō to maikyō 横川における如法
写経と埋経.‖ Kōkogaku zasshi 考古学雑誌 54 (1969): 1-30.
Kajiyama, Yuichi. ―Stūpas, the Mother of Buddhas, and Dharma-body.‖ In New Paths in
Buddhist Research, edited by A. K. Warder, 9-16. Durham, North Carolina: The Acorn
Press, 1985.
Kameda Tsutomu 亀田孜. ―Jūbun saishōōkyō jūkai hōtō mandara 重文最勝王経十界宝塔曼荼
羅.‖ In Chūsonji, edited by Ishida Mosaku, 68. Tokyo: Asahi Shinbunsha, 1959.
Kamens, Edward, trans. The Three Jewels A Study and Translation of Minamoto Tamenori’s
Sanboe. Ann Arbor: University of Michigan, 1988.
Kawakatsu Kenryō 川勝賢亮, ed. Tahōtō to Hokekyō shisō 多宝塔と法華経思想. Tokyo:
Tōkyōdō, 1984.
Keikai 景戒. ―Nihon Ryōiki 日本霊異記.‖ In Nihon koten bungaku zenshū 日本古典文学全集,
vol. 6, edited by Nakada Norio 中田祝夫. Tokyo: Shōgakkan, 1975.
Kidder, Jr., J. Edward. ―Busshari and Fukuzō: Buddhist Relics and Hidden Repositories of
Hōryūji.‖ Japanese Journal of Religious Studies 19 (1992): 217-244.
Kieschnick, John. ―Blood Writing in Chinese Buddhism.‖ Journal of the International
Association of Buddhist Studies 23, no. 2 (2000): 177-94.
Komatsu Shigemi 小松茂美. ―Menashikyō to sono shūhen 目なし経とその周辺.‖ Museum 60
(1956): 24-26.
______. ―Hokekyō sasshi ni tsuite 法華経冊子について.‖ Museum 81 (1957): 7-12.
______. ―Ichiji sanrei no shakyō 一字三礼の写経.‖ Museum 186 (1966): 2-11.
______. Heike nōkyō no kenkyū 平家納経の研究, vol. 1 and 2. Tokyo: Kōdansha, 1976.
Kujō Kanezane 九条兼実. Gyokuyō 玉葉, vol. 1. Tokyo: Sumiya Shobō, 1966.
Kuno Takeshi 久野健 and Hajime Nakamura, eds. Bukkyō bijutsu jiten 仏教美術事典. Tokyo:
Tokyo Shoseki, 2002.
Kuroita Katsumi 黑板勝美, ed.. ―Hyakurenshō 百錬抄.‖ In Shintei zōho, Kokushi taikei, vol. 11.
Tokyo: Kokushi Taikei Kankōkai, 1929.
______, ed. ―Honchō monshū 本朝文集.‖ In Shintei zōho, Kokushi taikei 新訂増補, 國史大系,
vol. 30. Tokyo: Kokushi Taikei Kankōkai, 1929.
______, ed. ―Bunji no chūmon 文治の注文.‖ In Azuma kagami 吾妻鏡. In Shintei zōho, kokushi
taikei 新訂增補, 國史大系, vol. 32, 352-55. Tokyo: Yoshikawa Kōbunkan, 1929-64.

284
Kyoto National Museum 京都囯立博物館, ed. Kana no bi かなの美. Tokyo: Ōtsuka Kōgeisha,
1996.
______. Ōchō no butsuga to girei zen o tsukushi bi o tsukusu 王朝の佛画と儀礼: 善をつくし
美をつくす. Kyoto: Kyoto National Museum, 1998.
______. Koshakyō seinaru moji no sekai 古写経: 聖なる文字の世界. Kyoto: Kyoto National
Museum, 2004.
______. Nichiren to hokke no meihō hanahiraku Kyōto machishū bunka 日蓮と法華の名宝: 華
ひらく京都町衆文化. Kyoto: Kyoto National Museum and Nihon Keizai Shinbunsha,
2009.
______. ―Ichiji rendai hokekyō.‖ Accessed August 7, 2011.
http://www.kyohaku.go.jp/jp/syuzou/index.html.
______. ―Konkōmyōkyō.‖ Accessed August 7, 2011.
http://www.kyohaku.go.jp/jp/syuzou/index.html.
LaFleur, William R. The Karma of Words: Buddhism and the Literary Arts in Medieval Japan.
Berkeley and Los Angeles: University of California Press, 1986.
Lai, Whalen W. ―The Predocetic ‗Finite Buddhakāya‘ in the Lotus Sūtra: In Search of the
Illusive Dharmakāya Therein.‖ The Journal of the American Academy of Religion XLIX
(1981): 447-469.
Lancaster, Lewis R. ―An Early Mahāyāna Sermon about the Body of the Buddha and the Making
of Images.‖ Artibus Asiae 36 (1974): 287-91.
______. ―The Oldest Mahāyāna Sūtra: Its Significance for the Study of Buddhist Development.‖
Eastern Buddhist 8 (1975): 30-41.
Legge, James. A Record of Buddhist Kingdoms: Being an Account by the Chinese Monk Fa-hien
of His Travels in India and Ceylon in Search of the Buddhist Books of Discipline. New
York: Dover, 1965[1886].
Lévi-Strauss, Claude. The Savage Mind. London: Weidenfeld and Nicolson, 1962.
Levine, Gregory. Daitokuji: The Visual Cultures of a Zen Monastery. Seattle: University of
Washington Press, 2005.
Lopez, Donald S. ―Inscribing the Bodhisattva's Speech: On the ‗Heart Sūtra's‘ Mantra.‖ History
of Religions 29 (1990): 351-72.
______. Elaborations on Emptiness: Uses of the Heart Sūtra. Princeton: Princeton University
Press, 1996.
Machida Seishi 町田誠之. ―Mukashi no shōsokukyō 昔の消息経.‖ Nihon rekishi 日本歴史
584 (1997): 31-33.
Makransky, John J. "Controversy over Dharmakāya in India and Tibet: A Reappraisal of Its
Basis, Abhisamayālaṃkāra Chapter 8." The Journal of the International Association of
Buddhist Studies 12, no. 2 (1989): 45-78.
______. Buddhahood Embodied: Sources of Controversy in India and Tibet. Albany, NY: State
University of New York Press, 1997.
Martin, Henri-Jean. The History and Power of Writing. Translated by Lydia G. Cochrane.
Chicago and London: University of Chicago Press, 1994.
Meech-Pekarik, Julia. ―Disguised Scripts and Hidden Poems in an Illustrated Heian Sutra:
Ashide and Uta-e in the Heike Nōgyō.‖ Archives of Asian Art 31 (1977): 52-78.
Mitchell, W.T.J. ―Word and Image.‖ In Critical Terms for Art History, edited by Robert S.
Nelson and Richard Shiff, 51-61. Chicago: University of Chicago Press, 2003.

285
Miya Tsugio 宮次男. Kinji hōtō mandara 金字宝塔曼荼羅. Tokyo: Yoshikawa Kōbunkan,
1976.
______. ―Kenrantaru kyōten 絢爛たる経典.‖ In Kenrantaru kyōten 絢爛たる経典, edited by
Sato Shinji, 90-124. Tokyo: Heibonsha, 1983.
______. ―Myōhōjizō myōhōrengekyō kinji hōtō mandara ni tsuite 妙法寺蔵妙法蓮華経金字宝
塔曼陀羅について.‖ Bijutsu kenkyū 美術研究 337 (1987): 88-96.
Mochizuki Shinkō 望月信亨. Mochizuki bukkyō daijiten, vol. 5. Tokyo: Sekai Seiten Kankō
Kyōkai, 1966-68.
Moerman, D. Max. ―The Archeology of Anxiety: An Underground History of Heian Religion.‖
In Heian Japan: Centers and Peripheries, edited by Mikael Adolphson, Edwards
Kamens, and Stacie Matsumoto, 245-71. Honolulu: University of Hawai‘i, 2007.
Nagao Gadjin. ―On the Theory of Buddha-body: Buddha-kāya.‖ Translated by Hirano Umeyo.
Eastern Buddhist 6 (1973): 25-53.
Nagata Mizu 永田瑞. ―Ryūju no busshinkan 竜樹の仏身観.‖ Indogaku bukkyōgaku kenkyū 印
度学仏教学研究 22 (1973): 369-72.
Nagazumi Yasuaki 永積安明 and Shimada Isao 島田勇雄, eds. Hōgen monogatari; Heiji
monogatari 保元物語; 平治物語. In Nihon koten bungaku taikei 日本古典文學大系,
vol. 31. Tokyo: Iwanami Shoten, 1961.
Nakamura Hajime 中村元. Iwanami bukkyō jiten 岩波仏教辞典. Tokyo: Iwanami Shoten, 1989.
Nakamura, Kyoko Motomochi, trans. Miraculous Stories from the Japanese Buddhist Tradition
The Nihon Ryōiki of the Monk Kyōkai. Cambridge, MA: Harvard University Press, 1973.
Nakayama Tadamichi 中山忠親. Sankaiki 山槐記. In Zōhō shiryō taisei 増補史料大成,
vol. 27. Tokyo: Rinsen Shoten, 1965.
Nara National Museum 奈良国立博物館, ed. Narachō shakyō 奈良朝写経. Tokyo: Tokyo
Bijutsu, 1983.
______, ed. Hokekyō shakyō to sōgon 法華経: 写経と荘厳. Tokyo: Tokyo Bijutsu, 1988.
______. Kamakura bukkyō kōsō to sono bijutsu 鎌倉仏教: 高僧とその美術. Nara: Nara
National Museum, 1993.
______, ed. Bukkyō setsuwa no bijutsu 仏教説話の美術. Kyoto: Shibunkaku, 1996.
______. Josei to bukkyō inori to hohomie 女性と仏教: いのりとほほみえ. Nara: Nara
National Museum, 2003.
______, ed. Danzan Jinja no meihō yamato no kamigami to bijutsu 談山神社の名宝: 大和の
神々と美術. Nara: Nara National Museum, 2004.
______, ed. Seichi Ninpō (Ninpō) Nihon bukkyō 1 00-nen no genryū 聖地寧波 (ニンポー): 日
本仏教 1300 年の源流. Nara: Nara Kokuritsu Hakubutsukan, 2009.
National Museum of Korea, ed. Pulsari changom. Seoul: National Museum of Korea, 1991.
National Palace Museum 國立故宮博物院. Gugong shuhuaji 故宮書画録, volume 8. Taipei:
National Palace Museum, 1965.
NHK Promotion プロモーション, ed. Yoshitsune ten Genji, Heishi, Ōshū Fujiwara-shi no
shihō 義経展: 源氏, 平氏, 奥州藤原氏の至宝. Tokyo: NHK Promotion, 2005.
Nishiguchi, Junko. ―Where the Bones Go: Death and Burial of Women of the Heian High
Aristocracy.‖ Translated by Mimi Hall Yiengruksawan. In Engendering Faith: Women
and Buddhism in Premodern Japan, edited by Barbara Ruch, 428-35. Ann Arbor: Center
for Japanese Studies, University of Michigan, 2002.
286
Ocho Enichi 横超慧日. ―Tahōtō shisō no kigen 多宝塔思想の起源.‖ Indogaku bukkyōgaku
kenkyū 印度学仏教学研究 2, no. 1 (1953): 30-36.
Ohnuma, Reiko. ―The Gift of the Body and the Gift of the Dharma.‖ History of Religions 37
(1998): 323-59.
______. Head, Eyes, Flesh, and Blood: Giving Away the Body in Indian Buddhist Literature.
New York: Columbia University Press, 2007.
Okada Yukihiro 岡田行弘. ―Nagarunja to buttō suiha ナーガールジュナと仏塔崇拝.‖
Indogaku bukkyōgaku kenkyū 印度学仏教学研究 48 (2000): 1076-71 [reverse
pagination].
Ōsaka Shiritsu Bijutsukan 大阪市立美術館, ed. Shōtoku Taishi shinkō no bijutsu 聖徳太子信仰
の美術. Ōsaka: Tōhō Shuppan, 1996.
Ōta Masahiro 太田正弘. ―Kokerakyō ni tsuite: Aichiken Iwakurashi Taisanji shutsudohin o
chūshin to shite 杮経について:愛知県岩倉市大山寺出土品を中心として.‖
Kokugakuin zasshi 国学院雑誌 76 (1975): 31-40.
Ōyama Jinkai 大山仁快. ―Shakyō 写経.‖ Nihon no bijutsu 日本の美術 156 (1979).
Payne, Richard K. ―Awakening and Language: Indic Theories of Language in the Background of
Japanese Esoteric Buddhism.‖ In Discourse and Ideology in Medieval Japanese
Buddhism, edited by Richard R. Payne and Taigen Dan Leighton, 79-96. London and
New York: Routledge, 2006.
Peirce, Charles Sanders. Collected Papers (1931-58), vol. 2. Edited by Charles Hartshorne and
Paul Weiss. Cambridge: Harvard University Press, 1932.
Rambelli, Fabio. ―Re-inscribing Mandala: Semiotic Operations on a Word and Its Object.‖
Studies in Central and East Asian Religions 4 (1992): 1-24.
______. ―The Semiotic Articulation of Hosshin Seppō: An Interpretive Study of the Concepts of
Mon and Monji in Kūkai's Mikkyō.‖ In Esoteric Buddhism in Japan, edited by Ian Astley,
17-36. Copenhagen and Aarhus: The Seminar for Buddhist Studies, 1994.
______. ―Texts, Talismans, and Jewels: The Reikiki and the Performativity of Sacred Texts in
Medieval Japan.‖ In Discourse and Ideology in Medieval Japanese Buddhism, edited by
Richard Karl Payne and Taigen Dan Leighton, 52-78. New York: Routledge, 2006.
______. Buddhist Materiality: A Cultural History of Objects in Japanese Buddhism. Stanford:
Stanford University Press, 2007.
Reischauer, Edwin O. Ennin's Diary: The Record of a Pilgrimage to China in Search of the Law.
New York: Ronald Press, 1955.
Roberts, Allen F. and Mary Nooter Roberts. A Saint in the City: Sufi Arts of Urban Senegal. Los
Angeles: UCLA Fowler Museum of Cultural History, 2003.
Robson, James. ―Signs of Power: Talismanic Writing in Chinese Buddhism.‖ History of
Religions 48 (2008): 130-69
Roth, Gustav. ―Symbolism of the Buddhist Stūpa according to the Tibetan version of the Caitya-
vibhāga-vinayodbhāva-sūtra, the Sanskrit treatise Stūpa-lakṣaṇa-kārikā-vivecana, and a
corresponding passage in Kuladattas Kriyāsaṃgraha.‖ in The Stūpa. Its Religious,
Historical and Architectural Significance, edited by Anna Libera Dallapiccola in
collaboration with Stephanie Zingel-Avé Lallemant, 183-209. Wiesbaden: Steiner, 1980.
Ruppert, Brian D. Jewel in the Ashes: Buddha Relics and Power in Early Medieval Japan.
Cambridge, MA: Harvard University Asia Center, 2000.

287
Sainen 西念. ―Sō Sainen konshi kinji kuyō nichiroku 僧西念紺紙金字供養日録.‖ In Insei jidai
no kuyō mokuroku 院政時代の供養目録, edited by Miyake Yonekichi 三宅米吉 and
Tsuda Noritake 津田敬武, 12-41. Tokyo: Tokyo Teishitsu Hakubutsukan, 1924.
Saitō Enshin 斎藤円真, trans. Jikaku daishi den 慈覚大師伝. Tokyo: Sankibō Busshorin, 1992.
Salomon, Richard and Gregory Schopen. ―The Indravarman (Avaca) Casket Inscription
Reconsidered: Further Evidence for Canonical Passages in Buddhist Inscriptions.‖ The
Journal of the International Association of Buddhist Studies 7, no. 1 (1984): 107-123.
Sasaki Hōsei 佐々木邦世. ―Kingin kōsho no tejun to kōfu 金銀交書の手順と工夫.‖ In
Kenrantaru kyōten 絢爛たる経典, edited by Sato Shinji, 132-34. Tokyo: Heibonsha,
1983.
Sasaki Kōkan 佐々木宏幹. ―Sō no jushika to ō no saishika: bukkyō to ōsei to no musubitsuki ni
kansuru ichi shiron 僧の呪師化と王の祭司化: 仏教と王制との結びつきに関する一
試論.‖ In Kokka to tennō tennōsei ideorogi to shite no bukkyō 国家と天皇 天皇制イデ
オロギ-としての仏教, edited by Kuroda Toshio 黒田俊雄, 49-91. Tokyo: Shujūsha,
1987
Satō Akio 佐藤亮雄, ed. Hyakuza hōdan kikigakishō 百座法談聞書抄. Tokyo: Nan‘undō
Ofūsha, 1963.
Satō Shinji, ed. Kenrantaru kyōten 絢爛たる経典. Tokyo: Heibonsha, 1983.
de Saussure, Ferdinand. Course in General Linguistics. Translated by Roy Harris. London:
Duckworth, 1983.
Schopen, Gregory. ―The Phrase ‗Sa Pŗthivīpradeśaś Caityabhūto Bhavet‘ in the Vajracchedikā:
Notes on the Cult of the Book in Mahāyāna.‖ Indo-Iranian Journal 17 (1975): 147-81.
______. ―Burial Ad Sanctos and the Physical Presence of the Buddha in Early Indian Buddhism.‖
Religion 17 (1987): 193-225.
Seki Hideo 関秀夫. ―Kyōzuka to sonno ibutsu 経塚とその遺物.‖ Nihon no bijutsu 日本の美術
292 (1990).
______. Heian jidai no maikyō to shakyō 平安時代の埋経と写経, 4th ed. Tokyo: Tōkyōdō
Shuppan, 1999.
Sekine Daisen 関根大仙. Mainōkyō no kenkyuū 埋納経の研究. Tokyo: Ryūbunkan, 1968.
Sharf, Robert H. "The Idolization of Enlightenment: On the Mummification of Ch'an Masters in
Medieval China." History of Religions 32, no. 1 (1992): 1-31.
______. "On the Allure of Buddhist Relics." Representations 66 (1999): 75-99.
______. ―Introduction: Prolegomenon to the Study of Japanese Buddhist Icon.‖ In Living
Images: Japanese Buddhist Icons in Context, edited by Robert H. Sharf and Elizabeth
Horton Sharf, 1-18. Stanford: Stanford University Press, 2001.
______. Coming to Terms with Chinese Buddhism: A Reading of the Treasure Store House.
Honolulu: University of Hawai‘i, 2002.
Shen, Hsueh-man. ―Realizing the Buddha‘s Dharma Body During the Mofa Period: A Study of
Liao Buddhist Relic Deposits.‖ Artibus Asiae 61, no. 2 (2001): 263-303.
Shibuya Kuritsu Shōto Bijutsukan 渋谷区立松濤美術館, ed. Mojie to emoji no keifu 文字絵と
絵文字の系譜. Tokyo: Shibuya Kuritsu Shōto Bijutsukan, 1996.
Shimatani Hiroyuki 島谷弘幸. ―Sōshokukyō no hassei to tenkai 装飾経の発生と展開.‖ In Shin
no bi shakyō no kokoro 信の美: 写経のこころ, edited by Chūsonji, 18-29. Hiraizumi:
Chūsonji, 2000.

288
Shimizu Masumi 清水眞澄. ―Nōdoku to nōsetsu: ongei ‗dokyō‘ no ryōiki to tenkai 能読と能説:
音芸‗読経‘の領域と展開.‖ Ryōjin kenkyū to shiryō 梁塵: 研究と資料 15 (1997): 25-
29.
______. Dokyō no sekai nōdoku no tanjō 読経の世界—能読の誕生. Tokyo: Yoshikawa
Kōbunkan, 2001.
Shinmura Izuru 新村出, ed. Kōjien 広辞苑, 4th ed. Tokyo: Iwanami Shoten, 1991.
Silbergeld, Jerome and Dora C.Y. Ching, eds. Persistence/Transformation: Text as Image in the
Art of Xu Bing. Princeton: P.Y. and Kinmay W. Tang Center for East Asian Art in
Association with Princeton University Press, 2006.
Snellgrove, David L. ―Śākyamuni's Final ‗nirvāṇa.‘‖ Bulletin of the School of Oriental and
African Studies 36, no. 2 (1973): 399-411.
______, ed. The Image of the Buddha. London: Serindia Publications/UNESCO, 1978.
Stevenson, Daniel B. ―Buddhist Practice and the Lotus Sūtra in China.‖ In Readings of the Lotus
Sūtra, edited by Stephen F. Teiser and Jacqueline I. Stone, 132-50. New York: Columbia
University Press, 2009.
Stone, Jacqueline I. ―Medieval Tendai Hongaku Thought and the New Kamakura Buddhism: A
Reconsideration.‖ Japanese Journal of Religious Studies 22 (1995): 17-48.
______. ―Chanting the August Title of the Lotus Sutra: Daimoku Practices in Classical and
Medieval Japan.‖ In Re-Visioning "Kamakura" Buddhism, edited by Richard Karl Payne,
116-66. Honolulu: University of Hawai'i Press, 1998.
______. Original Enlightenment and the Transformation of Medieval Japanese Buddhism.
Honolulu: University of Hawai‘i Press, 1999.
______. ―‗Not Mere Written Words‘: Perspectives on the Language of the Lotus Sūtra in
Medieval Japan.‖ In Discourse and Ideology in Medieval Japanese Buddhism, edited by
Richard R. Payne and Taigen Dan Leighton, 160-94. London and New York: Routledge,
2006.
Strong, John S. Relics of the Buddha. Princeton: Princeton University Press, 2004.
Sudō Hirotoshi 須藤弘敏 and Iwasa Mitsuharu 岩佐光晴. Chūsonji to Mōtsūji 中尊寺と毛越寺.
Tokyo: Hoikusha 保育社, 1989.
Suguro Shinjō 勝呂信静. ―Hokekyō no budda-ron 法華経の仏陀論.‖ In Hokke Bukkyō no
butsuda ron to shujō ron 法華仏教の仏陀論と衆生論, edited by Watanabe Hōyō 渡辺
宝陽, 61-110. Kyoto: Heirakuji Shoten 平楽寺書店, 1985.
Takahashi Tomio 高橋富雄. ―Chūsonji to hokekyō: Chūsonji konryū no kokoro 中尊寺と法華
経: 中尊寺建立の心.‖ Tōhoku daigaku kyōyōbu kiyō 東北大学教養部紀要 33 (1981):
19-41.
Takakusu Junjirō 高楠順次郎 and Watanabe Kaigyoku 渡邊海旭, eds. Taishō daizōkyō 大正大
藏経. Tokyo: Taishō issaikyō kankōkai, 1924-32.
Takasaki Jikidō 高崎直道. ―Dharmatā, Dharmadhātu, Dharmakāya and Buddhadhātu: Structure
of the Ultimate Value in Mahāyāna Buddhism.‖ Indogaku bukkyōgaku kenkyū 印度学仏
教学研究 14 (1966): 919-903 [reverse pagination].
Takeda Nichikatsu 竹田日濶. ―Nichiren no mandara to kihonzon tono kankei ni tsukete 日蓮の
曼荼羅と其本尊との関係に就て.‖ Indogaku bukkyōgaku kenkyū 印度學佛教學研究 6
(1958): 152-53.

289
Tambiah, Stanley Jeyaraja. Culture, Thought, and Social Action: An Anthropological Perspective.
Cambridge: Harvard University Press, 1985.
Tanabe, Willa J. Paintings of the Lotus Sutra. New York, Tokyo: Weatherhill, 1988.
Tanaka Kaidō 田中塊堂. Shakyō nyūmon 寫經入門. Ōsaka: Sōgensha, 1971.
______. Nihon shakyō sokan 日本寫經綜鑒. Kyoto: Shibunkaku, 1974.
Tōji 東寺, ed. Tōji no bijutsu kaiga to kōgei 東寺の美術: 絵画と工芸. Kyoto: Tōji, 1976.
Tokyo National Museum 東京国立博物館. Kin to gin: kagayaki no nihonbijutsu 金と銀: かが
やきの日本美術. Tokyo: Tokyo National Museum, 1999.
______. Kokuhō daigoji ten 国宝醍醐寺展. Tokyo National Museum, 2001.
______. Sho no shiho nihon to chūgoku 書の至宝: 日本と中国. Tokyo National Museum, 2006.
______. ―Konkōmyōkyō Volume Four Menashikyō.‖ Accessed August 7, 2011.
http://www.tnm.jp/modules/r_collection/index.php?controller=dtl&colid=B2400&t=type
_s&id=24&lang=ja.
Tsuji Zennosuke 辻善之助. Nihon no bukkyō shi 日本佛敎史, vol. 1. Tokyo: Iwanami Shoten,
1960.
Umehara Sueji 梅原末治. ―Kankoku keishū kōfukuji tō hakken shari yōki 韓国慶州皇福寺塔発
見の舎利容器.‖ Bijutsu kenkyū 美術研究 156 (1950): 31-47.
Ury, Marian. Tales of Times Now Past: Sixty-Two Stories from a Medieval Japanese Collection.
Ann Arbor: Center for Japanese Studies, University of Michigan, 1993.
Veidlinger, Daniel M. Spreading the Dhamma: Writing, Orality, and Textual Transmission in
Buddhist Northern Thailand. Honolulu: University of Hawai‘i Press, 2007.
Wallis, Glenn. ―The Buddha‘s Remains: Mantra in the Mañjusrimulakalpa.‖ Journal of the
International Association of Buddhist Studies 24, no. 1 (2001): 89-125.
Wang Ying and Yan Sun. Reinventing Tradition in a New World: The Arts of Gu Wenda, Wang
Mansheng, Xu Bing and Zhang Hongtu. Gettysburg: Gettysburg College, 2004.
Weidner, Marsha, ed. Flowering in the Shadows: Women in the History of Chinese and Japanese
Painting. Honolulu: University of Hawai‘i Press, 1990.
Woodward, F.L., trans. The Book of Kindred Sayings (Sanyutta-Nikāya) or Grouped Suttas, 5
vols. London: Oxford University Press, 1917-30.
Xu Bing. ―An Artist‘s View.‖ In Persistence/Transformation: Text as Image in the Art of Xu
Bing, edited by Jerome Silbergeld and Dora C.Y. Ching. Princeton: P.Y. and Kinmay W.
Tang Center for East Asian Art in Association with Princeton University Press, 2006.
Yajima Kyōsuke 矢島恭介. ―Konshi kinjikyō to busshari 紺紙金字経と仏舎利.‖ Museum 81
(1957): 2-6.
Yamakawa Kumiko 山川公見子. ―Kyōzuka no shinkō 経塚の信仰.‖ Kikan kōkogaku 季刊考古
学 97 (2006): 70-73.
Yamanaka Yutaka 山中裕, trans. Eiga monogatari 栄花物語, vol. 2. Tokyo: Shōgakkan, 1995-
98.
Yamato Bunkakan 大和文華館. Nezame monogatari emaki bungei to bukkyō shinkō ga orinasu
bi 寝覚物語絵巻: 文芸と仏教織りなす美. Nara: Yamato Bunkakan, 2001.
Yiengpruksawan, Mimi Hall. ―One Millionth of a Buddha: The Hyakumantō Darani in the
Scheide Library.‖ The Princeton University Library Chronicle 48 (1987): 224-38.
______. Hiraizumi: Buddhist Art and Regional Politics in Twelfth-Century Japan. Cambridge:
Harvard University Asia Center, 1998.

290
______. ―Illuminating the Illuminator: Notes on a Votive Transcription of the Supreme Scripture
of Golden Light (Konkōmyō saishō ōkyō).‖ Versus 83/84 (2000): 113-20.

291